Shorter Contributions to General Geology 1964 v GEOLOGICAL SURVEY PROFESSIONAL PAPER 504 ~ '3 T/zz's volume was pué/z'saea’ as separate caapters A—F except coméz'nea’ c/zapz‘ers Daaa’E UNITED STATES DEPARTMENT OF THE INTERIOR STEWART L. UDALL, Secretary GEOLOGICAL SURVEY Thomas B. Nolan, Director ,/ EARTH SCIENCES LIBRARY CONTENTS [Letters designate the separately published chapters] (A) Glacial reconnaissance of Sequoia National Park, California, by Francois E. Matthes. (B) Postglacial drainage evolution and stream geometry in the Ontonagon area, Michi- gan, by John ’1‘. Hack. (C) Geology and petrogenesis of the Island Park caldera of rhyolite and basalt, eastern Idaho, by Warren Hamilton. (D, E) Studies of the zeolites—Composition of zeolites of the natrolite group and Com- positional relations among thomsonites, gonnardites, and natrolites» by Margaret D. Foster. (F) Underground temperatures and heat flow in the East Tintic district, Utah, by T. S. Lovering and H. T. Morris. 94'? Q5 75 Pt v. ’50 LP EARTH SCIENCES LIBRARY . " GLACIAL RECONNAISSANCE OF SEQUOIA NATIONAL PARK CALIFORNIA GLACIAL RECONNAISSANCE OF SEQUOIA NATIONAL PARK The largest and most perfectly formed avalanche chute in Sequoia National Park, viewed from the High Sierra Trail east of camp in Bearpaw Meadow. Like its smaller companion, this chute is carved in massive exfoliating granite and terminates at the brink of the glacial U-shaped canyon below. The downward narrowing of the chute is explained by the protection given to the lower part of the chute by a snow cone on the surface of the glacier which lay in the canyon. Glacial Reconnaissance of Sequoia National Park California By FRANCOIS E. MATTHES Prepared porthumouxly by FRITIOF F RYXELL from M atthex’ not“ and other your“: SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY GEOLOGICAL SURVEY PROFESSIONAL PAPER 504—A Caaracterz'yz‘z'cx aaa’ dz'strz'autioa of Me aacz'em‘ glaciers in Me most mat/zed}! national paré of Me Sierra N evaa’a UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1965 UNITED STATES DEPARTMENT OF THE INTERIOR STEWART L. UDALL, Secretary GEOLOGICAL SURVEY Thomas B. Nolan, Director For sale by the Superintendent of Documents, US. Government Printing Office Washington, DC. 20402 FOREWORD In 1905, the US. Geologlcal Survey assigned Francois E. Matthes to make a large-scale (1224,000) map of Yosemite Valley. The classic Yosemite Special map which resulted enhanced Matthes’ well-established reputation as a master topographer, but paradoxically it contributed to his decision to discontinue topographic surveying. His sojourn among the sublime but imperfectly understood features of the “Incomparable Valley” had confirmed a conviction that his deepest interest lay in studying the genesis of landforms rather than in depicting them on topographic maps. Matthes’ decision in favor of geology was not the result of any sudden impulse. He had long been a serious student of glacial geomorphology and geology, he had published papers in these fields, and he had spent a year at Harvard University taking advanced courses under William Morris Davis, the “father of geomorphol- ogy.” The opportunity to engage in geologic work on a full-time basis came in 1913, in which year Matthes was officially transferred from the Topographic Branch of the Survey to the Geologic Branch. The Yosemite mapping project had other far-reaching consequences. Quite logically, the first geologic assignment given Matthes was to return to Yosemite to investigate the origin of this celebrated valley, a subject which had been a matter of controversy ever since the Whitney surveys in the 1860’s. Matthes’ painstaking research eventually produced USGS Professional Paper 160, “The Geologic History of Yosemite Valley,” a monograph of great and enduring significance. As his Yosemite investigations progressed, Matthes felt the need for testing his tentative conclusions by comparative study of other valleys in the Sierra Nevada that, presumably, had had much the same history. Therefore he extended his field studies into several major drainage basins of the west slope. These studies, besides serving their immediate purpose, opened up many new geological vistas, some of which involved fundamental questions pertaining to the Sierra Nevada as a whole. After the Yosemite report was published, Matthes gave renewed attention to these broader problems and during the 1930’s he made a succession of geological reconnaissances, particularly in the central and southern Sierra Nevada. On these explorations he employed whatever means were practicable. For the most part, he depended on pack trips, a mode of transportation at which he had become expert from many years of topographic surveying. “Packing in” enabled him to establish camps in remote regions otherwise inaccessible because they were far from railroads, roads, and even trails. Areas surrounding his base camps were explored on horseback or on foot. In the later years, when better roads were available, he got about also by car. On one occasion, in 1936, he employed a chartered plane to make an aerial survey of the High Sierra in the Mount Whitney region. (His companion on this . trip was Conrad Wirth, later Director of the National Park Service.) The difficulties of the Sierra terrain make Matthes’ accomplishments in covering the territory all the more impressive. There were few parts of the Sierra Nevada, especially throughout the middle and upper reaches of the broad west slope, which he did not come to know firsthand. The times demanded much such reconnaissance work, and a number of geologists became highly proficient at it. Matthes was outstandingly successful in making geological surveys in mountainous regions. The results of Matthes’ Sierra studies were summarized in several general papers. How- ever, he earnestly desired to report more fully on certain specific areas. This hope was never realized. Though he worked on several longer papers whenever he could do so, the oppor- tunities were infrequent, for other assignments of more immediate urgency if not greater importance generally took precedence. Thus he came to regard the completion of these reports as something that would have to wait until after his retirement. On June 30, 1947, VI FOREWORD he did retire, and he moved to California; but the following year, before he was fairly started with his writing, he suffered a brief and fatal illness. Matthes had entered into negotiations with the University of California Press for the writing of two books intended for the general public. In 1949, at the request of the pub- lisher and Mrs. Matthes, I endeavored to carry out these plans; and consequently, in 1950, two works were published (Matthes, 1950a, b) in connection with observance of the Centen- nial Year of California. In the Geological Survey and in the Sierra Club, interest was also manifested in posthumous works which might embody the substance of Matthes’ studies, even though they could not attain the exact form or completeness that he would have given them. The Geological Survey assigned me to look into these possibilities. Review of the Matthes’ papers indicated that some of the materials left by Matthes might best serve for reference use, in connection with future glacial and geomorphological studies, but that two significant reports might be prepared for publication. There resulted, as a first product of this undertaking, USGS Professional Paper 329, “Reconnaissance of the Geomorphology and Glacial Geology of the San Joaquin Basin, Sierra Nevada, California,” published in 1960. The paper herewith published is the second of the two reports planned. In a sense these are companion works. However, since Matthes’ geomorphological studies in Sequoia National Park have previously been published (Matthes, 1937, 1938, 1950a) and little new information could be added, the emphasis in this paper is on glacial geology, and only brief incidental attention is given to such subjects as erosional history, exfoliation, and avalanche sculpture. The glacial terminology that Matthes used is retained in this report, although it departs from the current Geological Survey usage as outlined in the Code of Stratigraphic Nomen- clature prepared by the American Commission on Stratigraphic Nomenclature (1961). For example, Matthes used the term “stage” for glacial episodes such as El Portal, Glacier Point, and Tahoe, whereas the code (p. 660, art. 40) specifies the term “glaciation” for these episodes. Similarly, Matthes used the term “substage” for subdivisions of a glaciation, whereas current usage employs the term “stade.” A rather detailed description of the topography of Sequoia National Park (“Geographic Sketch”), which Matthes had almost completed, is included in full because of the valuable picture is gives of the park area. As a preliminary to the preparation of this report, I spent the period from July 22 to August 27, 1952, in Sequoia National Park working from camps in Giant Forest, Bearpaw Meadow, Mineral King, and the Kern Canyon. This fieldwork refreshed my memory of features already familiar and acquainted me with others not previously seen. Subsequently, a procedure was followed that was, in the main, similar to that adopted for preparation of Professional Paper 329. The attempt was made to prepare a unified account that would be as complete as possible from the materials at hand, and that would set forth Matthes’ observations and conclusions, and retain, wherever feasible, his own words. This procedure involved the synthesis of information from many different sources, particularly field notes and maps, sections of text written in longhand with various degrees of finish, annotated photographs, and published papers. The project further necessitated both the transfer of information from field maps to an unfinished ofiice map and the completion of three topo- grahic profiles. No outline for a report was found, but the source materials at hand, as well as those used for the San Joaquin report, provided working guides for the organization of material and filling in of gaps in the report. The photographs proved particularly useful. In taking these, Matthes had kept in mind the needs of the interpretative program of Sequoia National Park and had made a‘ special effort to secure a comprehensive, illustrative series. Many of these he had annotated fully for use in his so-called “Sequoia Albums” (Matthes, 1950a). It is hoped that this publication will in some measure fulfill Matthes’ intent to provide an overall picture of the geography and glacial geology of Sequoia National Park, and also, that it may prove useful not only to geographers and geologists, but to others, such as the CONTENTS Page Glaciation of the Kaweah Basin—Continued Page Foreword __________________________________________ V East Fork _______________________________________ A34 Abstract ___________________________________________ A1 South Fork _____________________________________ 34 Introduction ........................................ 2 Glaciation of the upper Kern Basin: Kern glacier system- 35 Geographic sketch _________________________________ 2 Kern Canyon and tributary valleys _______________ 36 Kaweah Basin ——————————————————— . ——————————————— 5 Wisconsin Stage _____________________________ 37 Upper Kern Basin _______________________________ 7 Kern Canyon ___________________________ 37 Kern Canyon ——————————————————————————————————— 12 Tributary valleys _______________________ 38 Glaciation .......................................... 17 E1 Portal Stage _____________________________ 45 Basis for differentiation of the glacial stages _________ 19 Glacier Point Stage _________________________ 52 Wisconsin Stage _____________________________ 20 Summary statement _________________________ 53 El Portal Stage _____________________________ 22 Local glaciation south of Sequoia National Park Glacier Point Stage _________________________ 24 (west of main Sierra crest) ..................... 54 Glaciation of the Kaweah Basin: Kaweah glacier system- 26 sources of the Little Kern River _______________ 54 Marble Fork ___________________________________ 28 Sources 0f the Tule River ____________________ 54 Wisconsin Stage _____________________________ 28 Sources of Golden Trout Creek and South Fork El Portal Stage _____________________________ 29 of the Kern River _________________________ 54 Glacier Point Stage _________________________ 30 Selected references __________________________________ 55 Middle Fork ___________________________________ 30 Index ______________________________________________ 57 ILLUSTRATIONS [Plates are in pocket] FRONTISPIECE. Photograph of the largest and most perfectly formed avalanche chute in Sequoia National Park, viewed from the High Sierra Trail east of camp in Bearpaw Meadow. PLATE 1. Map of ancient glaciers of Sequoia National Park. 2. Profiles through Sequoia National Park. Page FIGURE 1. Index map ______________________________________________________________________________________ A3 2—3. Idealized section across the Sierra Nevada— 2- Representing the simple “textbook conception” of its tilted-block structure ______________________ 4 3. In the latitude of Sequoia National Park ____________________________________________________ 4 4—7. Photographs: 4. View eastward toward headwaters of Middle Fork, Kaweah River ______________________________ 5 5. Panoramic View from Alta Peak of the broad western slope of the Sierra Nevada _________________ 6 ' 6. View up unglaoiated lower part of Kaweah River canyon, toward the platform on which the Giant Forest stands and, at the right, Moro Rock ________________________________________________ 6 7. View westward from Moro Rock, along the cliffs bordering the platform on which the Giant Forest 7 stands ________________________________________________________________________________ 8. Simplified profile across the upper Kern Basin, showing remnants of four ancient landscapes (erosion surfaces) at different levels above the Kern Canyon ________________________________________________________ 8 IX 748—960 O—65——2 X CONTENTS FIGURES 9—51. Photographs: 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 32. 33. 34. 35. 36. . Kern River, cutting in bedrock, near Kern Canyon Ranger Station at south border of Sequoia 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. Westward across the Kern Canyon to the Chagoopa Plateau; Kaweah Peaks Ridge in the back— ground ________________________________________________________________________________ Northwestward from the main Sierra crest across the upper Kern Basin to the Great Western Divide- Southeastward toward cirques on the Great Western Divide, which forms part of the boundary between Sequoia and Kings Canyon National Parks ______________________________________ Northward from Mount Whitney toward Mount Russell (center) ______________________________ One of the many lakelets occupying glacially quarried rock basins in the upper Kern basin, above the junction of Milestone Creek __________________________________________________________ Northwestward from Mount Guyot _________________________________________________________ Southward down the Kern Canyon, from a point on the west rim near mouth of Rattlesnake Creek- - Southward down the Kern Canyon, from a point below the rim, north of Wallace Creek __________ Chagoopa Falls, which descends the steep west wall of the Kern Canyon from a small hanging valley on the Chagoopa Plateau ________________________________________________________________ Slotlike gulch incised in the east wall of Kern Canyon by a streamlet descending from the mouth of a hanging valley below Kern Lake _________________________________________________________ . Kern Canyon from a point about 2 miles south of the park boundary ___________________________ . One of the timbered moraines of the Wisconsin Stage that surround Moraine Lake _______________ . A frost-split block of granite on one of the Wisconsin moraines that encircle Moraine Lake ________ . Disrupted glacial boulder in Wallace Canyon ________________________________________________ . Glacier polish, striae, and grooves, above the head of Kern Canyon _____________________________ . Fantastic rock forms in the upper Kern Basin, near Lake South America ________________________ . Across the glaciated floor of Whitney Canyon, showing combined effects of quarrying and grinding-- . A 16-foot erratic left on Bighorn Plateau by the ice of El Portal Stage __________________________ . Glacial boulder of El Portal Stage, resting on a platform overlooking the Big Arroyo ------------- . Glaciated knob at the head of South Fork, Kaweah River _____________________________________ . Saddle east of Tower Rock, on the east rim of Kern Canyon ___________________________________ . Alta Peak (11,211 feet) ___________________________________________________________________ . Outwash of El Portal Stage, revealed in a roadcut on the Generals Highway, about half a mile above Ash Mountain Park Headquarters ________________________________________________________ Glacial outwash, dating perhaps from the Glacier Point Stage, exposed in a roadcut on the Generals Highway above Camp Potwisha __________________________________________________________ “River Valley,” the glaciated upper canyon of Middle Fork, Kaweah River _____________________ Spectacular summits south of Hamilton Lakes ____________________________________ ‘ ___________ Near the source of Hamilton Creek, Iboking down the canyon across one of the Hamilton Lakes-___ Sand Meadows ___________________________________________________________________________ National Park _________________________________________________________________________ Vicinity of the Nine Lake Basin southward down Big Arroyo, which, in the distance (left), becomes a deep, U-shaped canyon flanked by forested plateaus ______________________________________ Across Big Arroyo from an unnamed mountain east of Little Clair Lake ________________________ Chagoopa Plateau and Moraine Lake, viewed from the edge of Big Arroyo ______________________ Moraine Lake, on the Chagoopa Plateau ____________________________________________________ Up the valley of Tyndall Creek, from the Bighorn Plateau that appears in the lower right foreground“ Northeastward up the valley of Wright Creek toward Mount Tyndall (central peak with gullied slopes) ________________________________________________________________________________ Tulainyo Lake, high on the main crest of the Sierra Nevada; viewed from the west _______________ Gorge 50 feet deep, cut by Wallace Creek, for the most part in postglacial time __________________ Mount Whitney, viewed from the west ______________________________________________________ Junction of Crabtree Canyon (foreground) and Whitney Canyon (background) __________________ Whitney Trail across upper Whitney Canyon and Hitchcock Lake (foreground) at Mount Hitchcock (center) _______________________________________________________________________________ North side of Mount Hitchcock, viewed across Whitney Canyon _______________________________ East face of Mount Hitchcock _____________________________________________________________ ' Eastward across the Kern Canyon in the vicinity of Kern Canyon Ranger Station ________________ Page A9 10 11 12 13 13 14 14 16 16 17 20 21 22 23 24 25 26 26 27 27 29 30 31 31 32 33 35 38 39 40 40 41 42 43 44 45 46 47 48 49 50 51 SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY GLACIAL RECONNAISSANCE OF SEQUOIA NATIONAL PARK, CALIFORNIA By FRANQOIS E. MATTHES AB STRACT Sequoia National Park, like Yosemite and Kings Canyon National Parks to the north, is on the broad west slope of the Sierra Nevada. It extends from an elevation of about 1,400 feet in the western foothills to 14,495 feet at Mount Whitney, the culminating summit on the main crest of the range at the east. Thus its altitudinal range, about 13,100 feet, is greater than that of any other national park in the United States, south of Alaska. The park embraces the most southerly portion of the High Sierra, the scenic higher part of the range. The main crest of the Sierra Nevada, at the east border of the park, bears many high peaks, no less than seven of which have altitudes exceeding 14,000 feet. Traversing the central part of the park from north to south is a secondary crest, the Great Western Divide, which is likewise an impressive range with peaks 11,000 feet to over 13,000 feet in altitude. This crest divides the park into two approximately even but dissimilar halves. The western half is occupied by the basin of the Kaweah River, most southerly of the southwestward-flowing master streams of the Sierra Nevada. The Kaweah Basin is an intricately dissected, rugged area of high relief. The eastern half is occupied by the upper basin of the Kern River. The Kern is unique among the master streams in that it flows directly southward, nearly parallel to the main crest of the range. Distinctive features of the upper Kern Basin, in addi- tion to the high bordering mountain crests, are the impressive U-shaped Kern Canyon and the broad benches (ancient erosion surfaces) that border this canyon and its branches. In the Pleistocene Epoch both the Kaweah Basin and the upper Kern Basin were occupied by glacier systems. These were the most southerly of the major glacier systems of the Cascades-Sierra Nevada chain. Being less favorably situated than those to the north, they were of smaller volume; neverthe- less, glaciers of considerable size formed in both basins, espe- cially the Kern, during each of the three glacial stages—the Glacier Point Stage, the El Portal Stage, and the Wisconsin Stage. Information concerning the characteristics and distri- bution of the glaciers was sought by distinguishing and mapping the morainal deposits of each of the stages. In the Kaweah Basin, the development of glaciers was limited by the fact that this basin heads not along the lofty main crest of the Sierra Nevada but on the Great Western Divide and on other secondary crests that are only part way up the Sierra west slope. Evidence is present in this basin for the earliest stage, the Glacier Point, but is extremely meager. For the next stage, El Portal, and the most recent stage, the Wisconsin, the records are far better. They indicate that during both of these stages the converging canyons of the Kaweah Basin became pathways for cascading ice streams. Even the larger of these streams, however, attained a length of only 10 miles. The ice streams therefore fell short of uniting to form a major trunk glacier corresponding to the ones in the main drainage basins to the north and in the Kern Basin to the east. There formed, instead, relatively small separate glacier systems, one or more in the headwater areas of each of the main branches of the Kaweah. Only locally, in their upper reaches, did these glaciers oversweep the divides; for the most part the glaciers were confined to the canyons, and these they filled only in part. The lowest altitude reached by ice in El Portal Stage was about 4,550 feet; in the Wisconsin Stage, about 5,200 feet. The Kern glacier system, by contrast, was a great, many- branched ice body fed from ranks of cirques along the high bordering ranges. Since the Kern Canyon extends in a nearly straight line through the middle of the upper Kern Basin, and the tributary canyons branch from it like the ribs in an oak leaf, the Kern glacier system had much the same leaflike pat- tern. The maximum extent reached by the Kern glacier system in the Glacier Point Stage cannot be determined with certainty, but the evidence would seem to warrant the inference that the glacier advanced approximately as far as its successor of El Portal Stage. Records of El Portal Stage, though incomplete, can be inter- preted with more assurance. The volume of ice was then greater in some places than the canyons could hold; the ice locally spread across intervening divides and over benchlands on either side of the main canyon to a total breadth of 4 to 6 miles, thus producing a central ice sheet about 30 square miles in extent. The overall length of the Kern glacier system was 32 miles; the terminus of the trunk glacier lay at an altitude of 5,700 feet in the bend of the canyon to the north of Hockett Peak (at lat 36°14’, which may represent the southern limit reached by glacial ice in the Sierra Nevada). Records of the Wisconsin Stage are for the most part very well preserved. They indicate that during this stage the Kern glacier system had less volume than during El Portal Stage and remained a sprawling ice body whose trunk and branches lay confined within their respective canyons as distinct ice streams, separated from one another by mountain spurs or low divides. The tributary glaciers were as much as 15 miles long. The overall length of the Kern glacier system was 25 miles. The farthest point reached by the terminus of the trunk glacier coin- cides with the south boundary of the park; the boundary posts stand at an altitude of 6,350 feet on the curving outer moraine that marks the extreme limits of the Wisconsin glaciation. The time available for this reconnaissance did not permit subdividing the Wisconsin Stage into substages. However, A1 A2 morainal complexes were noted that may throw light on this subject as detailed investigations are undertaken in the future. The more significant postglacial changes in the park include those which have served to modify the form of the Kern Canyon. The present U—shaped form of that canyon is not precisely the one that resulted from repeated glacial erosion. The walls, once smooth, are now furrowed by gullies, and talus slopes at their base define the curves of a new U-shaped form superimposed on the glacially eroded one. The low gradient of the canyon over a stretch of several miles suggests that the glaciated rock floor may have been excavated into a chain of lake basins that are now filled with sediments. INTRODUCTION The field data upon which this report is based were gathered in the course of three field seasons devoted wholly or in part to a reconnaissance of Sequoia Na- tional Park for the US. Geological Survey. The in- itial fieldwork was done in 1925 between June 9 and July 15. For the next decade the author’s research was diverted to other sections of the Sierra Nevada, but in 1935 he resumed work on Sequoia National Park—- this time at the request of the National Park Service and with the effective cooperation of that agency—with a view to completing a systematic reconnaissance. In that year he spent the period from July 8 to December 18 in the field, and in 1936, during the interval from May 27 to July 15, he concluded the project. Lawson (1904) was the first to give a comprehensive sketch of the geomorphology and glacial geology of Sequoia National Park, and his classic report provided an excellent insight into the mode of development of the major landforms of that region through successive cycles of erosion and through glaciation. Some details on the glaciation of the Mineral King region were pro- vided by Knopf and Thelan (1905). But the need for more complete information on the genesis of the landscape of the entire park—for the scenic features of this park are so exceptional as to be secondary only to the “Big Trees” which give it name—led to the joint sponsorship of the reconnaissance reported in this and previous papers (Matthes, 1937, 1938, 1947, 1950a). Necessarily, because of the mountainous character of the tract, the physical difficulties which it offers to travel—especially through its wilder parts—and, also, because of the shortness of the working season in the High Sierra portion, much of which is above the tim- berline, the survey could be only of a preliminary nature. It will readily be understood that the inter- pretations which resulted are in part tentative and in need of verification by future, more intensive studies. Nor did the reconnaissance cover all the different aspects of the geology with equal thoroughness. Par— ticular attention was given to the evolution of the land- forms, and relatively little time was devoted to the SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY examination of the different types of rock that occur within the area. There are so many different forma- tions, and these are so complexly related to one another, that to map them individually and study the petrologic character 'of each of them would alone have required several field seasons. Some of the results of the re- connaissance have already been published, particularly those pertaining to the geomorphological features. In the present report, additional results are set forth, especially With reference to the character and distri- bution of the ancient glaciers which occupied Sequoia National Park during the various stages of the Pleisto— cene Epoch. The author wishes to thank officials of the National Park Service for the support they gave this survey and for their warm interest in it. While the fieldwork was in progress, staff members of Sequoia National Park were generous with their time and assistance on many occasions and in many ways. Thanks are in order to several individuals who provided illustrations which appear in this report (all photographs not taken by‘the author have been credited to those who furnished them). The University of California Press freely gave permis— sion to quote extensively from its publication, “Sequoia National Park: a Geological Album” (Matthes, 1950a). Finally, grateful appreciation is expressed to Clyde A. Wahrhaftig for critical reading of the manuscript and for the constructive suggestions that served to improve its effectiveness. GEOGRAPHIC SKETCH Sequoia National Park (fig. 1) comprises an irregu- lar tract 604 square miles in extent on the west slope of the Sierra Nevada between lat. 36°18’ N. and 36°42’ N. and between long. 118°14’ W. and 119°00’ W. (pl. 1). It embraces chiefly the scenic upper part of the range, including the main crestline, and in that respect it is situated much like Kings Canyon National Park, which adjoins it directly on the north, and Yosemite National Park, which lies a hundred miles farther to the north- west (figs. 2, 3). But Sequoia National Park has the distinction of including the culminating stretch of the crestline, bearing Mount Whitney (14,495 ft) and six more of the eleven 14,000-foot peaks of the Sierra Nevada; the park has the additional distinction of being traversed from north to south by a secondary crest that is but little lower than the main crest—the Great West- ern Divide. Noteworthy also is the fact that Sequoia National Park has the greatest range of altitude of any national park or national monument in the United States, south of Alaska. Its lowest point, in the canyon of the Kaweah River near the entrance below Ash Mountain Park Headquarters, is only about 1,400 feet GLACIAL RECONNAISSANCE 0F SEQUOIA NATIONAL PARK 9 «5° SACRAME T0 4/ . Mojave o A O N O O 8 . :5 ”X 5 ”war! < 5’ N“ D River V‘ a ° . g 0,581 Orovflle tr fi“ Yllba R1219,- / RIfiEF\ ’ss/EQUO \ A \ ,P R & ALABAMA HILLS OWENS LAKE 40 MILES | . 'YOSEMITEX .P'\;,r I‘ NA IONA PAfiKl/L °’ §\ NATIONfi ‘ 06 \. as % l w \ XNATIONAI} /' y ‘\ , \ / , \ AJMONO LAKE\ :——-" \\ \ .‘/\ . . A; giaderaf/ (MONUMENT, i SAN / IN \ /fJOAQ . FIGURE 1.-—Index map showing location of Sequoia National Park. A3 A4 SHORTER San Joaquin Valley 7* CONTRIBUTIONS T0 GENERAL GEOLOGY Mount Dana FIGURE 2.—Idealized section across the Sierra Nevada representing the simple “textbook con- ception” of its tilted-block structure. That conception fits approximately the facts as they \are known in the central part of the range, in the latitude of Yosemite Valley. Both northward and southward, however, the structure becomes more complex. San Joaquin Valley Mount Whitney Alabama Hills Owens FIGURE 3.—Idealized section across the Sierra Nevada in the latitude of Sequoia National Park. Step faults exist at the western margin of the range as well as at the eastern margin, and the west slope consequently breaks off rather abruptly in the foothills belt. Some distance from the foothills, the tops of peaks on the down—faulted buried fault block emerge as isolated rocky hills above the sedimenw that fill the San Joaquin Valley. From Matthes (1950a, p. 3). above sea level. Thence to the top of Mount Whitney, therefore, the change in altitude amounts, in round num- bers, to 13,100 feet. All of the three national parks in the Sierra Nevada include parts of that scenic upper region near the main crestline which Californians aptly call the “High Sierra.” This region is a mountain land of truly alpine character, whose jagged snow—flecked peaks rise high above the timberline and whose strongly glaciated val- leys are dotted with hundreds of picturesque lakes and lakelets. Of this alpine upper country, Sequoia Na- tional Park embraces the southernmost part. Its southern boundary, indeed, coincides approximately with the line where the High Sierra comes to an end; thence southward the range assumes a more subdued aspect, none of its summits rising above timberline. In one other respect Sequoia National Park differs markedly from its two sister parks. Both Yosemite and Kings Canyon National Parks are traversed by southwestward—trending rivers—that is, by rivers that flow directly down the west slope of the range, roughly parallel to the rank and file of the master streams. The drainage net of Sequoia National Park, on the other hand, is complicated by the secondary mountain crest previously mentioned, the Great Western Divide. To the west of that divide is the drainage basin of the Kaweah River, the southernmost of the great series of southwestward-flowing master streams. To the east of the divide, on the other hand, is the headwaters basin of the Kern River; this river flows directly southward in a nearly straight line, parallel to the main crestline, and maintains that unusual course for 75 miles before turning in a southwesterly direction toward the Great Valley of California. The Kern River, moreover, heads against the South Fork of the Kings River, to the north, and as a consequence the Kaweah River does not reach up to the main crestline. The Great Western Divide is far more than a mere ridge or crest; it is a mountain range in itself (fig. 4). Surmounted by a row of sharp-profiled peaks from 11,000 to more than 13,000 feet in altitude, and un- broken by any profound gaps,1 it constitutes in effect a formidable barrier that trends from north to south across the park and divides it into two approximately even but very dissimilar halves. The divide stands 2,000 to 3,000 feet above the valleys at its eastern base and 4,000 to 6,000 feet above the canyons at its western base. Viewed from points in the western half of the park—as from More Rock, on the edge of the Giant F orest—the Great Western Divide has the appearance of a spectacular alpine range, and as a consequence it is commonly mistaken by sightseers for the main crest of the Sierra Nevada. The main crest actually lies behind the Great Divide, completely hidden from view. Finally, a word about the country between Sequoia National Park and the western foothills of the Sierra Nevada. In an airline, that country measures 12 to 15 1The only trail-passes over the Great Western Divide are Kaweah Gap (10,700 ft), Black Rock Pass (11,500 ft.), Franklin Pass (11,300 ft.), Shotgun Pass (11,300 ft), and Coyote Pass (10,034 ft). GLACIAL RECONNAISSANCE OF SEQUOIA NATIONAL PARK A5 FIGURE 4.——View eastward toward headwaters of Middle Fork, Kaweah River. The snow-clad peaks on the skyline are on the Great Western Divide. In the left foreground is More Rock, an imperfect dome that owes its rounded form to long- miles in width. From the point where the Kaweah River leaves the park to the embayment in which it debouches upon the plain of the San Joaquin Valley, the airline distance is only 11 miles. Looking down the Kaweah River canyon from the top of Alta Peak or from the rim of the Giant Forest platform, one is im— pressed by the multiplicity of rugged ranges and spurs advancing toward the canyon from either side, one behind another, at successively lower heights, parallel to the border of the distant plain. The name that has been given to one of the viewpoints on the Generals Highway—Eleven Range Point—well expresses the effect which the landscape makes upon the spectator. Between these successive ranges and ridges are deep- cut canyons and gulches tributary to the Kaweah. The majority of these canyons trend at right angles to the southwesterly course of the master stream, but an in- spection of the topographic maps shows that this trend is by no means universal and that, taken as a whole, the drainage pattern is extremely varied and irregular. Nor is the alinement of the ridges with the foot of the range as parallel and as persistent as their appearance in the distant View would seem to indicate. The map shows, furthermore, that. not all are sharp crested; a continued exfoliation of the massive granite. At the lower right is the canyon of the Middle Fork. Aerial photograph by Frank Webb. considerable proportion have broad undulating summit areas. There is no complete gradation of summit levels all the way down to the San Joaquin Valley. That fact is evident at once to the traveler who approaches the moun- tain range from the west, by way of either Lemon Cove or Woodlake. Even the first outlying hills which he passes rise abruptly 500 to 1,000 feet above the level of the plain, and the first continuous ridges of the moun- tain mass rise 1,500 to 2,000 feet above the plain. These ridges form an irregular, steep mountain front that resembles an escarpment. KAWEAH BASIN The western half of the park is profoundly dissected by the converging branch canyons of the Kaweah River system (fig. 5). Though none of its peaks rise to great altitude, it is an extremely rugged piece of country, diffi— cult to traverse save on manmade roads or trails. The main canyon of the Kaweah River at Ash Mountain Park Headquarters is 3,000 to more than 4,000 feet deep. Ash Peak (5,621 ft), on the northwest side, stands 4,100 feet above the river, and Milk Ranch Peak (6,305 ft), on the southeast side, 4,800 feet. The canyons of the A6 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY FIGURE 5.—Panoramic view from Alta Peak of the broad west slope of the Sierra Nevada. The Kaweah River canyon, left of center, is one of the many deep trenches cut by south- westward-flowing rivers. The foothills are 27 miles away. Beyond them is the level San Joaquin Valley, deeply filled North, Middle, East, and South Forks are even deeper. Paradise Peak (9,370 ft) stands 5,400 feet above the East Fork and 6,400 feet above the Middle 'Fork. Homers Nose (9,005 ft) stands 5,200 feet above Crlough Cave, on the South Fork, and 5,500 feet above the East Fork. The Castle Rocks (9,150 ft) rise 6,150 feet above the Middle Fork, and Alta Peak (11,211 ft) rises 6,910 feet above it. Some parts of the Kaweah Basin, however, are not with silt, sand, and gravel washed down from the bordering ranges. On the horizon, barely discernible because of the haze, are the Coast Ranges, more than a hundred miles dis- tant. Photograph by L. Moe. deeply dissected by canyons but consist of undulating plateaulike uplands traversed by shallow valleys. Of these upland areas—which are clearly surfaces record- ing successive stages in the erosional history and rise of the Sierra Nevada (Matthes, 1930, 1933, 1937, 1950a, 1960)—the most extensive is in the headwaters of the Marble Fork. The platform on which the Giant Forest stands (figs. 6, 7 ), at altitudes ranging from 6,500 to 7,000 feet, is a characteristic part of this upland; Lodge- FIGURE 6.——View up unglaciated lower part of Kaweah River canyon toward the platform on which the Giant Forest stands. to the fact that it is composed of massive granite. At the right, Moro Rock. More Rock owes its prominence in the landscape The platform is also held up mainly by massive granite, but the mountain slopes below have been eroded from normally jointed rocks, partly granitic, partly metamorphic. GLACIAL RECONNAISSANCE 0F SEQUOM NATIONAL PARK A7 FIGURE 7.—View westward from More Rock along the cliffs bordering the platform on which the Giant Forest stands. Part of the Giant Forest is seen at the right. The cliffs are of sparsely jointed granite that is exfoliating very slowly and rather irregularly. Similar massive granite outcrops else- pole and Tokopah Valleys, in the middle course of the Marble Fork, also form part of it. High above the Giant Forest, however, on the summit of Panther Peak (9,044 ft) is another much smaller platform or flat; 2,000 feet above that, again, is a third flat on the top of Alta Peak (11,211 ft). Thus the landscape appears to rise by successive stories, each being marked by a level of gently undulating surface that contrasts with the steep mountain slopes between. UPPER KERN BASIN The upper Kern Basin, which forms approximately the eastern half of the park, contrasts strikingly with the Kaweah Basin, both in general configuration and in arrangement of drainage lines (pl. 2). Instead of being intricately dissected by a maze of branching can- yons and gulches, the Kern Basin has a broadly open, spacious aspect. This spaciousness is due not merely to the fact that the neighboring mountain ranges—the Great Western Divide on the west and the main crest of the Sierra Nevada on the east—stand 13 to 18 miles apart but also to the presence of extensive terracelike benchlands that flank the central canyon and give the basin What appears, at least from a distance, to be a broad and nearly level floor. Those benchlands, like those in the Kaweah Basin, are ancient erosion surfaces, 748—960 0—65—33 where on the platform indicate that the whole platform is made up largely of this durable material; it is no doubt to this circumstance that the platform, a remnant of an ancient erosion surface, owes its preservation. better preserved here, perhaps, than in any other section of the Sierra Nevada. From these benchlands and from the even older erosion surfaces recognizable in the tab— ular summits of certain of the peaks may be read the record of the rise of the Sierra Nevada. Their charac- teristics and significance were set forth in the pioneer study of Lawson (1904) and in the later studies of Webb (1946) and of the author (Matthes, 1930, 1933, 1937, 1950a,1960). (See fig. 8.) At its head the Kern Basin is encircled by the Kings- Kern Divide and the Great Western Divide, which form one continuous jagged mountain range. The Great Western Divide is strongly bowed toward the west, but the main crest of the range is locally bowed towards the east; the basin as a whole is therefore spoon shaped in outline. Even more apt, in view of the simple drainage pattern, is its likeness to a foliage leaf, as suggested by Lawson (1904). The straight Kern River traverses the basin axially like the midrib of the leaf, and the tributary streams branch from it at intervals, like veins. This simple pattern, it should be observed, is limited strictly to the upper basin, which lies within the limits of the park. Farther south, the drainage net becomes more complex, some of the tributary streams running parallel to the southflowing Kern River for consider- able distances. Examples are the Little Kern River, A8 Kaweah Peaks Mount Kaweah SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY Mount Langley Cirque 14,042 13, 752 13,8/5 13,451 peak 2,863 Boreal Plateau 11,500 Chagoopa Plateau FIGURE 8.—Simplified profile across the Upper Kern Basin, showing remnants of four ancient landscapes (erosion surfaces) at different levels above the Kern Canyon. The gently sloping summit of Mount Langley, like that of Mount Whitney, is a remnant of a hill in the lowland (Whitney erosion surface) that existed before the first major uplift of the Sierra region took place. The rounding summit of Cirque Peak is representative of a more mountainous landscape (Cirque Peak erosion surface) formed during the long interval between the second uplift and the third. The Boreal Plateau is a large remnant of an undulating landscape (Boreal Plateau erosion surface) that was also produced during the interval between the second uplift and the third. Chagoopa Plateau and the correspond- ing benches on the east side of Kern Canyon are remnants of a broad valley (Chagoopa erosion surface) that was evolved during the interval between the third uplift and the fourth. The Kern Canyon has been cut since the fourth and latest uplift, which took place at about the beginning of the glacial epoch. It is a product of alternate stream erosion and glacial erosion and is still in the process of being cut deeper. From Matthes (1950 a, p. 12, 13). Vertical exaggeration x 2. which flows southward for a distance of 16 miles be- tween the southern portion of the Great Western Divide and the Hockett Meadows Plateau, and the South Fork of the Kern, which pursues an irregular but, in the main, southerly course for some 50 miles (air-line dis- tance) through the eastern part of the Sierra Nevada before turning southwestward to join the master stream. The upper Kern Basin, further, is much less deeply trenched than is the Kaweah Basin. Throughout most of its length the Kern Canyon is only 2,000 to 2,500 feet deep. Whereas the Kaweah River has cut its canyon down to an altitude of 1,400 feet at the point where it leaves the park, the Kern River has an altitude of 6,400 feet where it crosses the southern park bound- ary. It follows that the upper Kern Basin, as a Whole, is a region of relatively great altitude. The broad benchlands flanking the Kern Canyon range from an altitude of about 8,000 feet at the southern boundary of the park to more than 11,000 feet at the head of the basin. The benchlands vary greatly in width—from a few yards to several miles—owing to the presence of bold spurs and mountain groups that project here and there from the enclosing ranges. At intervals, moreover, these benchlands are cut across by tributary canyons: yet, as is evident at once in any comprehensive View, they are remarkably persistent throughout the length of the basin. Though by no means level—from the flanking mountain ranges they slope down to brinks of the Kern Canyon at rates of 500 feet and more to the mile, and, in addition, are interrupted by low ridges and vales—the benchlands together form a distinct story in the landscape that contrasts with the towering moun— tains above and the steep—sided canyon below. They form a story not unlike that of the Giant Forest plat- form or of the Hockett Meadows Plateau in the Kaweah Basin but far more extensive and more clean-cut than either of these. Particularly broad and typically developed is that section of benchland which is known as the Chagoopa Plateau (fig. 9). It is on the west side of the Kern Canyon immediately above the junction of the great side canyon called the Big Arroyo. This plateau ex- tends 3 miles back from the canyon rim and in that distance rises from about 8,500 feet to more than 10,500 feet in altitude. Immediately above it loom Mount Kaweah (13,816 ft) and the great Red Spur (13,186 ft) which, together with the Red Kaweah (13,754: ft) and the Black Kaweah (13,752 ft), form one of the most imposing mountain groups that are attached to the Great Western Divide. To the tourist public the Cha— goopa Plateau is well known, for it is traversed by the High Sierra Trail, the main horse trail leading from the Giant Forest to Mount Whitney. The Chagoopa Plateau and the benchlands of cor- responding altitude on the opposite side of the Kern Canyon are not, however, the only story in the land- scape of the upper Kern Basin. As in the Kaweah Basin, so here there are several plateaulike flats or benches at different levels one above another (pl. 2). A prominent example is the Boreal Plateau, a gently undulating upland bench averaging more than 11,000 feet in altitude and stretching for a distance of 7 miles along the southeastern boundary of the park. Above GLACIAL RECONNAISSANCE OF SEQUOIA NATIONAL PARK A9 FIGURE 9.—View westward across the Kern Canyon to the Chagoopa Plateau; Kaweah Peaks Ridge in the background. The timbered plateau on both sides of the canyon is a rem- it, again, is the broadly convex summit of Cirque Peak, which ranges from 12,200 to 12,863 feet in altitude (pl. 2). And high above that peak rises the summit plat- form of Mount Langley, which ranges from 13,800 to 14,042 feet in altitude. Even Mount Whitney, the high- est peak of the Sierra Nevada, has a broad, gently slop- ing summit platform (figs. 8, 10). The platform, which is closely analogous to the one on Mount Langley, rises from 14,000 to 14,495 feet in altitude. Indeed, so very similar in general form are the two peaks that in 1871 Clarence King (1872), one of the early explorers of the Sierra Nevada, mistook Mount Langley for Mount Whitney and climbed it thinking that he was ascending the culminating peak of the range. It is not to be inferred from the foregoing that flat plateaulike summits are the rule or even are prevalent on the main crest of the Sierra Nevada. The summits of Mount Whitney and Mount Langley, and the lower one on Cirque Peak, are the only ones on that part of the main crest that flanks the Kern Basin. Farther north in the range, a few flat summits occur at long intervals—for example, on Mount Darwin (13,841 ft and 13,701 ft) and on Mount Wallace (13,328 ft) at the head of the upper San Joaquin Basin (Matthes, 1960, p. 38, 41) and on a continuous platform (12,500 to 13,000 ft) 31/2 miles long in Yosemite National Park nant of an erosion surface; that forming the Chagoopa Pla— teau is particularly striking. In the foreground, at the lower right, is a small hanging valley. (Matthes, 1937, p. 8—9)—bu>t most of the peaks are sharp profiled, as are those on the Great Western Di— vide. Only one of the peaks on the Great \Vestern Di- vide, Table Mountain (13,646 ft), has a clean—cut tabu- lar summit platform (fig. 11), although a few others have ill-defined sloping summit areas. The Great Western Divide and the main crest of the range are much alike in general character, indeed, are much like the majority of the serrate mountain crests that traverse the High Sierra for more than a hundred miles northward. Each consists essentially of a single chain of lofty angular peaks connected with each other by narrow, sharp crested, often splintered or pinnacled ridges (fig. 12). Their sprawling spurs likewise are for the most part narrow and sharp crested, Whereas the intermediate canyons as a rule are broadly U-shaped and head in steep-walled amphitheaterlike bowls. These ranges accordingly possess predominantly attenu— ated, sharply pointed forms and in a sense are “skeleton ranges.” Some parts of the main crest of the range— in the vicinity of Mount Barnard, Mount Whitney, and Mount Langley—however, depart somewhat from that general character, the forms at these places being more fullbodied, the summits tabular or gently sloping, and the spurs broad enough to have rounding contours. In both ranges, however, the U-shaped canyons and A10 SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY FIGURE 10.—View northwestward from the main Sierra crest across the upper Kern Basin to the Great Western Divide. An ancient erosion surface is preserved on the summits of Mount Whitney, in the foreground, and Mount Young, the the capacious amphitheaters seem wholly out of pro- portion to the small volume of the streamlets that de- scend through them. These streamlets in many places find their way among the knolls and bumps of the rock floors without definitely incised channels—here cascad- ing over rock steps that as yet show no signs of stream erosion, and there meandering through grassy meadows or losing themselves in gemlike lakelets (fig. 13). These features stand out all the more vividly in the landscape because the canyons and amphitheaters lie mostly, or in some places entirely, above the timberline, in the lower fringes of what is properly termed the “Alpine zone.” From the lofty summits on the main crest of the Si— erra Nevada, one looks down its precipitous east front into Owens Valley below. Particularly impressive is the view from Mount Whitney, not merely because of the great height of that peak above the valley—40,800 feet, or more than 2 miles—but because the spectator stands at the immediate, sharp-cut brink of a precipice that falls away sheer 1,500 feet beneath his feet. 80 breathtaking is the scene that the entire mountain front seems to drops off sheer like a wall. Yet, as is clearly shown on the accurately constructed topographic map, adjacent peak. Other, younger erosion surfaces form the plateaus bordering the Kern Canyon. Numerous cirques scal- lop the main Sierra crest and the Great Western Divide. Aerial photograph by Roy Curtis. the foot of the range is fully 5 miles out from the peak in horizontal distance. Even more incredible does it seem that the base of the' great escarpment is itself fully 3,000 feet above the valley floor on which stands the town of Lone Pine. What appears to be a level plain at the foot of the range is in reality a slope that descends 2,000 feet in a distance of 5 miles. And what appear to be insignificant hillocks at the farther limits of the slope are the fantastically shaped crags and pinnacles of the Alabama Hills, a picturesque miniature range that looms 1,000 to 1,400 feet above Lone Pine and has furnished the weird set— ting for many a moving picture allegedly filmed in India within sight of the great Himalayas—this moun— tain system being adequately represented in the back— ground by the Sierra Nevada. Over the broad level floor of Owens Valley, one can readily trace the serpentine course of the Owens River, conspicuous by reason of its fringe of deep-green bushes. Beyond the river stretches the long—drawn somber-hued Inyo Range, which forms the east wall of the valley, parallel to the Sierra Nevada. From altitudes of 8,000 and 9,000 feet at its southern end, the Inyo Range £an Mamurm N3 HEEwBOHE 3.84m 6336 one we 835:5 wnssfiaoc 2: we 98 3 53:8 mo Ewi 2: 3 Mama Enofiwu a5 .5335? Easy .353 Enoflwz guano meM and fiosvmm $953 E3582 25 «0 fig mafia EEK .35me 5333 38w 3: no 8556 6.538 Ewaamaofinow ngléa $5th . A11 GLACIAL RE‘CONNAISSANCE OF SEQUOIA NATIONAL PARK A12 FIGURE 12.—View northward from Mount Whitney to- ward Mount Russell (center). Here, as a result of the headward quarrying of the glaciers that form- erly occupied the opposing cirques, there remains of the former main crest of the range only the rock wall that connects the two peaks. The abundance of loose rock waste shows that the granite here breaks up readily—more readily than avalanches, running water, and gravity can remove the debris. The destructive action of alternating frost and thaw doubtless is promoted by the numerous vertical joint fractures. Photograph by Kenneth Flewelling. rises irregularly to altitudes of 13,000 and even 14,000 feet in its northern part which, because of its contrast— ing light-colored rocks, is known as the White Moun- tains. Southeastward, diagonally across Owens Valley and over gleaming Owens Lake in its bosom, are the rela- tively low Coso Mountains and several more distant mountain groups, and, more distant still, the lofty Pan- amint Range, behind which, deeply ensconced, lies Death Valley. The stupendous panorama from the summit of Mount Whitney is in truth second only to the Big Trees among the park’s natural exhibits as an inspiration for wonder and thought. KERN CANYON The great Kern Canyon extends due south through Sequoia National Park and divides that part of the Kern River basin lying within the park into two nearly equal parts. SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY ' From a geomorphological point of View, the head of the canyon may be regarded as being at Junction Meadow, where three branch canyons unite (fig. 14). The central branch canyon, though alined with the Kern Canyon, is distinctly subordinate to it in size and is separated from it by a great canyon step; this branch canyon is therefore of the rank of a tributary to the main canyon rather than a part of it. In its 17-mi1e stretch from Junction Meadow to the Vicinity of Coyote Creek, where it leaves the park, Kern Canyon follows a nearly straight course that doubtless reflects structural control, as suggested by Lawson (1904) and further analyzed by Webb (1936, 1946), this control being a fault or zone of faulting. The canyon is a typical glacial U-shaped trough (figs. 15, 16). Indeed, it is no overstatement to say that few glacial canyons, either in the Sierra Nevada or else— where, possess so nearly perfect a U—shaped form due to glacial excavation and maintained for so long a distance. The principal reason for its regularity is that the trough is sunk along the axis of a geomorphically postmature valley of great breadth and consequently has sharply defined rims, or shoulders, formed by the intersection of its precipitous walls with gently sloping uplands at both sides. (These shoulders, it is true, are broken at intervals by side canyons, and in other places on the uplands they make way for hills of moderate height.) In the section within Sequoia National Park, the Kern Canyon is deepest at its head and becomes pro- gressively more shallow southward. At Junction Meadow it is 2,500 to 2,600 feet deep; at the mouth of the Big Arroyo, 10 miles farther downstream, it is 2,000 feet deep, and opposite the rocky knob 14 miles below Junction Meadow, it is only 1,600 feet deep. These depths, it may be objected, are measured from the rims of the flanking uplands, and those uplands are not a mathematical plane, but parts of an undulat— ing, locally even hilly erosion surface. Allowance therefore should be made for the inequalities in that surface. To obviate errors from this source, the three measurements cited above were made at localities where the flanking uplands slope gently toward the rims and most probably represent the marginal parts of the ancient valley in which the U-shaped trough is sunk. Another source of error cannot be avoided. At Junc~ tiou Meadow the U-shaped trough is 1% miles wide from rim to rim, whereas it narrows to about 1 mile in its lower half; the depth indicated at Junction Meadow is therefore likely to be excessive, for the walls of the U—shaped trough there cut higher into the sloping up— lands than they do in the narrow part of the canyon. Careful examination of the topographic map, however, RECONNAISSANCE GLACIAL OF SEQUOIA NATIONAL PARK FiGURE 13.—0ne of the many lakelets occupying glacially quar- ried rock basins in the upper Kern Basin, above the junction of Milestone Creek. The granite at the sides, being only The Kern Canyon extends from the center of the picture toward the lower left; three branch canyons unite at its head. Photograph by Carl F. J. Overhage. FIGURE 14.—View northwestward from Mount Guyot. shows that the error thus introduced probably does not amount to more than a few hundred feet. That being admitted, there can be no doubt that the progressive southward shallowing of the U-shaped trough is due primarily to the fact that its floor has sparsely fractured, was not readily quarried away and conse- quently shows the effects of abrasion. a decidedly lower grade than does the bordering up- land surface. The grade of the canyon floor is reliably represented on the topographic map, for a line of spirit levels was run by the Geological Survey up the bottom of the Kern Canyon from Grasshopper Meadow (41/2 miles south of the park boundary) to Junction Meadow. The first stretch of 7 miles above Coyote Creek—that is, up to the mouth of the Big Arroyo—has a very low grade. This is the stretch that is aggraded behind a morainal dam. The altitude of the bench mark at Coyote Creek is 6,458 feet; the altitude of the bench mark at the mouth of the Big Arroyo is 6,664 feet. The rise in that 7 —mile stretch, therefore, is only 206 feet, or, on an average, 29 feet per mile. From the Big Ar- royo to the mouth of Rock Creek, the grade steepens gradually to an average of more than 100 feet per mile; and between Rock Creek and Whitney Creek it reaches locally as high as 200 feet per mile. But from the fan of Whitney Creek to Junction Meadow, it again flattens to less than 100 feet per mile. In many of its geomorphic features, the Kern Can- yon, in Sequoia National Park, resembles the Kings A14 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY FIGURE 15.——View southward down the Kern Canyon from a point on the west rim near mouth of Rattlesnake Creek. The pronounced U-shaped form of the canyon has been evolved by glacial erosion from a narrow V-shaped trench. FIGURE 16.———View southward down the Kern Canyon from a point below the rim north of Wallace Creek. The canyon is not straight throughout, although its course is probably determined by a fault or by several closely spaced parallel faults. Photograph by Kenneth Flewelling. The walls, once smooth, are now furrowed by gullies; talus slopes at their base produce the curves of a new U-shaped form superimposed on the glacially eroded one. Photograph by L. Moe. Canyon (canyon of the South Fork of Kings River) which adjoins it on the north (Matthes, 1926). Kern Canyon lacks imposing cliffs of massive rock like those above Zumwalt Meadows in the Kings Canyon, but it is a longer and in some respects a more impressive canyon. Its clifls are composed of sparsely jointed granite, and along many of the vertical or nearly ver— tical master joints, storm waters have cut clefts, some fully 100 feet deep and with cavelike recesses. Near Junction Meadow, on the west side of the canyon, there are clefts, a hundred yards or more apart, which have been cut along oblique master joints dipping steeply southward. I The floor of the canyon, throughout this section, is so largely covered with surficial deposits that only in a very few places does the river flow over bare rock. Glacial moraines are confined to a small area at the park boundary, but alluvial and talus deposits are ex- tensive throughout this section of the canyon. Un- doubtedly some of the material mantling the bedrock floor of the canyon is glacial out-wash. GLACIAL RECONNAISSANCE 0F SEQUOIA NATIONAL PARK At places where tributary streams join the Kern ’ River, great boulder fans extend out from the walls of the canyon onto its floor. Such a fan, composed of mingled lava and granite boulders, occurs at the mouth of Golden Trout Creek, and across the can- yon is the exceptionally large fan built by Coyote Creek at a place where Wisconsin moraines descend to the canyon floor. At the mouth of Rock Creek, the canyon floor is covered with large quantities of bouldery mate- rial deposited both by Rock Creek and by the Kern River, but here a distinct fan is lacking. Whitney Creek has a conspicuous fan, and Wallace Creek has an even larger fan which includes much material de- rived from lateral moraines of the Kern trunk glacier higher on the valley side. Wallace Creek now flows in a steep—walled trench that is cut across the lateral moraines and down through the upper part of the fan. All the alluvial fans have been truncated by the river at a considerable height above the canyon floor proper. The resultant scarps, though commonly 50 to 60 feet high, vary in height, depending on how steeply the sur— faces of the fans slope and on the extent to which these fans have been trimmed back by the river. As a con- sequence, in some places a scarp on one side of the river may be twice as high as one on the opposite side. These features are well illustrated by the great fan of Coyote Creek, which is typically truncated by a main scarp 60 to 70 feet above the river. This scarp is old enough to have numerous short gulches worn in it, and the edge of the terrace is therefore distinctly lobate. The lower terrace is nearly 20 feet above the river. A steel suspension bridge is built at this level. Conterno’s old suspension bridge, half a mile farther upstream, is built in a less favorable place, where, because the lower terrace is absent on the east side, the trail had to be cut obliquely down the scarp of the main terrace to the bridge. The partial dismantling of the canyon walls has given rise to long talus slopes which extend along both sides of the canyon. Although these slopes are nearly con- tinuous, there are, nevertheless, many places where rock in its original position crops out near the base of the talus. The talus slopes, which help to complete the U-shaped form of the Kern Canyon ( although the can— yon would be distinctly U-shaped without them), extend to various heights; near Junction Meadow the slopes apparently reach about halfway up the canyon side, but their exact upper limits are difIiCult to judge because of the brush cover. A puzzling feature of the talus slopes is the fact that those on the east side of the can- yon are for long distances more voluminous than those on the west side. Like the alluvial fans, the talus slopes have been cut back, and scarped, by the river. 748—960 0—615—d—4 A15 In places the canyon floor is encumbered by enormous blocks that have fallen from the bordering cliffs. Many of these blocks, which constitute one of the impressive features of the canyon, can be seen along the trail near Funston Meadows. They measure 10 to at least 50 feet in diameter but, as a result of irregular spalling, most are now much smaller than when they fell. No true exfoliation was noted in the Kern Canyon, in striking contrast to the many exfoliating granite bould- ers in the Kings River and upper San Joaquin River basins. The reason for this difference is probably to be found in the occurrence of irregular minor structurw in the granite of the Kern Canyon. The rock is full of shear fractures, mostly chloritized, and it tends to spall along these fractures, at least for short distances. Where tributary streams leave the mouths of hang- ing valleys on the uplands and descend into the canyon, cascades and waterfalls are found. Several of these are of spectacular height and great beauty, yet are still (1964) unnamed. The most notable waterfalls occur where several streams draining Chagoopa Plateau plunge down the steep west wall of the canyon. Of this group the falls of Red Spur have the greatest drop, about 2,300 feet. Chagoopa Falls (fig. 17) descend by several deep plunges, glissades, and, farther down, broken cascades, a total fall of- about 1,400 feet. As a result of stream entrenchment, some of the cascades lie in the bottom of gulches; where stream cutting has been facilitated ‘by fractures in the rock, the gulches may be so narrow and deep that the cascades recessed in them are all but invisible (fig. 18, 51). Immediately south of the park, there occur features which have significant bearing on the glacial history of Sequoia National Park (fig. 19). In this area, extend— ing about 7 miles southward from the park boundary, the canyon retains its distinctive broad U-shape but loses some of its regularity and, departing from its nearly straight southward course, swings eastward around Hockett Peak. Within this part of the canyon are ancient moraines which record the southerly limits of the Kern trunk glacier at one of its early stages. The Kern Lakes, also in this section, are not of glacial origin, however, but came into existence, as Lawson (1904, p. 343—345) cor- rectly recognized, through two special and different causes. Kern Lake is impounded behind the fan of a streamlet coming down from a hanging valley on the east side of the canyon; Little Kern Lake, which is sep- arated from the river by an almost continuous natural levee, is held up behind a dam resulting from a colossal rockslide which fell from the east wall of the canyon. Of great interest also are the prominent buttresses, in the bottom of this part of the canyon, which Lawson A16 FIGURE 17,—Chagoopa Falls, which descends the steep west wall of the Kern Canyon from a small hanging valley on the Chagoopa Plateau. The side valley was left hanging primarily as the result of rapid trench- ing by the master stream, but its height was in- creased by glacial deepening of the Kern Canyon. Widening of the canyon by glacial erosion also steep- ened the descent of the cascades. Kenneth Flewelling. Photograph by (1904, p. 331—343) regarded as being a type of geomorphic form not previously recognized and for which he proposed the name of “kernbut.” Lawson’s description of the kernbuts follows, in part: A remarkable feature of the Upper Kern, below Volcano Creek, is the departure of the stream from the west wall of the cafion and its crowding upon the east wall. This displacement of the stream is due to obstructions in the shape of a series of rocky buttresses, which adhere to the foot of the west wall, and, projecting out beyond the middle line of the cafion, locally, con- strict it, causing the stream to occupy narrow gorges between these buttresses and the east wall. In the interval between these buttresses the bottom of the caflon has its normal width of about half a mile from wall to wall. There are several of these buttresses in the vicinity of the Kern Lakes, and the two lakes lie in two of the intervals. In cross profile these buttresses have the character of rather sharp-crested ridges which run parallel to the general trend of Kern Canon; and a buttress may be a single ridge or a series of two or three ridges, in which case the latter are successively lower in the east. The buttresses may, therefore, be distinguished as single or multiple according as they present one or more of these ridges in cross profile. SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY FIGURE 18.—Slotlike gulch incised in the east wall of Kern Canyon by a streamlet descending from the mouth of a hang- ing valley below Kern Lake. Stream cutting was facilitated here by a fracture in the granite. The cascade is now so deeply recessed as to be almost invisible. This gulch is just outside the south boundary of the park, but similar gulches are found farther north, Within the park. The correlative pass, or col, which intervenes between the kernbut and the main canyon wall or between the parallel ridges of a multiple kernbut, Lawson desig- nated as “kerncol.” The names “kernbut” and “kerncol” were chosen, in part, because they are purely descriptive and carry with them no implication as to the genesis of the forms. GLACIAL RECONNAISSANCE 0F SEQUOIA NATIONAL PARK A17 FIGURE 19.—View up the Kern Canyon from a point about 2 miles south of the park boundary. Kern Lake, in the fore— ground, is not a glacial lake; it was formed in 1867—68 when the Kern River became pended through rapid growth of an alluvial fan. Beyond the lake, in the shadow, appears one of the buttresses (kernbuts) characteristic of this part of the GLACIATION In the 1860’s, when Professor Josiah D. Whitney (State Geologist of California, for whom Mount VVhit- ney was named), sent the first scientific exploring par- ties into the Sierra Nevada, and in the 1870’s, when John Muir and Professor Whitney engaged in their memorable controversy about the glaciation of the Yose- mite Valley, the concept of the Ice Age, or glacial epoch, was still very new and ill defined. Sufficient evidence was at hand to show that the north-central and north— eastern parts of the United States and the adjoining parts of Canada had once been covered by a vast ice sheet, but the precise limits which that ice sheet had reached were not definitely known. It was assumed to have formed part of an immense icecap that centered at the North Pole and to have mantled all the northern half of North America. Not unnaturally, this sheet was supposed to have also covered the great western mountain belt of the continent, many of whose ranges still hear glaciers at the present time. Not until later In the Wisconsin Stage, the Kern glacier reached canyon. only to the park boundary, but in the earlier stages it ex- tended several miles farther south and occupied the part of the canyon shown in the foreground of this view. Photograph by W. L. Huber. did it become clear that no icecap could have been at the North Pole of the Earth, because that pole was covered by an ocean of considerable depth. The ice sheet, it was found, originated on the North American continent itself—in Labrador and the region to the west of Hudson Bay, from which centers the ice flowed in all directions. ‘ It was discovered, further, that this ice sheet did not overwhelm the western mountain ranges but, anomal- ously, stopped to the east of them, and that those ranges were independent centers of snow accumulation which generated large glaciers of their own. Those glaciers became confluent and filled the intermontane valleys with ice to depths of several thousand feet. A vast composite ice mass was thus formed that was not an icecap, strictly speaking, nor did it bury the higher peaks; yet it was continuous over the entire breadth of the mountain belt. This Cordilleran ice, as it is appropriately called, spilled eastward onto the plains and westward into the fiords of the Alaskan coast and A18 British Columbia. It lay almost wholly to the north of the Canadian boundary line but sent several broad lobes southward into Puget Sound, eastern Washington, and western Montana. South of this composite ice mass, the higher moun- tain ranges also generated glaciers. The Sierra-Cas- cade chain is notable in that it bore glaciers throughout most of its great length. This chain extends for 1,000 miles over 14° of latitude from the 49th down to the 35th parallel. Thus the chain traverses regions of the utmost geographic and climatic diversity. It begins near the Canadian boundary, in a region of extremely wet, snowy climate, and terminates at the edge of the Mohave Desert, which is one of the driest and most torrid areas on the North American Continent. The glacial covering of this chain was very extensive at the north end, where the great height and breadth of the Cascade Range, the enormous quantities of snow supplied by the westerly winds, and the prolonged win- ters together produced conditions favorable for glacia- tion. This part of the chain lay fairly smothered under snow and ice and sent forth glaciers 80 to 100 miles long. South of Mount Rainier, however, the ice mantle con- tracted rapidly in breadth, mainly as a result of declin- ing altitude. Throughout southern Washington and northern Oregon, where the crestline, not counting the isolated volcanic peaks, rises scarcely above 5,000 feet, the glaciers attained lengths of only a dozen or so miles. Still farther south, throughout the 200-mile stretch of which Crater Lake, Mount Shasta, and Lassen Peak are the dominant landmarks, there was no true ice mantle but only detached glaciers and snowfields that lay in sheltered canyons high up on the main peaks. This dearth of ice—this state of semiaridity—was due not to a further decline in altitude, for the range here again rises to 6,000 and in places even to 8,000 feet, but to deficient snowfall caused by the presence between the Cascade Range and the Pacific Ocean of a large com- plex of mountains which intercepted a considerable share of the moisture from the westerly winds. Southward from the canyon of the Feather River, however, in the northern part of the Sierra Nevada, glaciation formed on an increasingly large scale, owing both to greater altitude and to greater snowfall; here the intercepting power of the Coast Ranges diminished wit-h decline in height. In the vicinity of Lake Tahoe, where the Sierra Nevada attains altitudes of more than 9,000 feet, the icefields and ice streams were large enough to coalesce and produce trunk glaciers from 15 to 20 miles in length. And in the stretch from Lake Tahoe to Mount Lyell, in which the crest rises progressively to altitudes of 11,000, 12,000, and 13,000 feet, the snows SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY were so abundant as to mantle the range continuously over a breadth of 20 to 30 miles and to create trunk glaciers 40 to 60 miles long. In short, another climax of glaciation was reached in this central part of the Sierra Nevada—a climax second only to that attained near the Canadian boundary, 800 miles to the north. From Mount Lyell to Mount Whitney, over a stretch of fully 100 miles, the glacial mantle extended almost undiminished in breadth. It covered all those parts of the High Sierra which are drained by the San Joaquin, Kings, Kaweah, and Kern Rivers and which are crowned by the culminating peaks of the range. A short distance south of Mount Whitney, however, within the boundaries of Sequoia National Park, the 300—mile-long glacial mantle of the Sierra Nevada came abruptly to an end. The glacier system of the upper Kern Basin was the southernmost large system of its kind in the Sierra Nevada. Beyond this system there were only a few small detached ice bodies. The southern limit of glaciation in the Sierra Nevada was imposed not primarily by latitude—that is, by the southward increase in warmth and consequent rise of the snowline during Pleistocene times—but by the termi- nation of the two lofty mountain ranges that bound the upper Kern Basin, for only these two ranges were high enough to reach above the snowline during glacial times. The precise extent of the glacial covering of the Sierra Nevada was long a subject of conjecture and dis- pute, but as a result of the systematic survey of glacial deposits, the margin of the glacial mantle has been de- finitely mapped, and its position can now be determined tentatively within narrow limits. There is thus no fur- ther doubt that the glacial mantle was confined wholly to the upper parts of the Sierra and that at no point did it reach down to, or even near, the western base of the range. In the south-central part of the range, where the glacial mantle was broadest, its western margin descended to altitudes of somewhat less than 5,000 feet. The trunk glaciers, of course, descended to still lower levels, yet even these glaciers fell far short of reaching the foot of the range. The Tu'olumne glacier, which was the longest ice stream north of the Yosemite region, attained a maximum length of 60 miles and projected about a dozen miles beyond the margin of the ice mantle that lay on the adjoining uplands. The Tuolumne glacier terminated, however, fully 30 miles from the foot of the range and at an altitude of about 2,000 feet. The Yosemite glacier at the time of maximum glaciation was 37 miles long and projected 7 miles beyond the margin of the ice mantle. The terminus of this glacier lay in the Merced Canyon about 50 miles from the foot of the range and about 2,000 feet above it, just below the site of _'T‘ ,g_~,v:,_ GLACIAL RECONNAISSANCE 0F SEQUOIA NATIONAL PARK El Portal. The San Joaquin glacier was nearly as long as the Tuolumne glacier, but it advanced only a few miles beyond the margin of the ice mantle on the flank- ing uplands. The San Joaquin glacier halted 45 miles from the mouth of its canyon, at an altitude of about 2,600 feet. The Kings glacier, most southerly of the great glaciers, despite the great altitude of the crest re— gion which it drained, attained a length of only 44 miles (measured along its middle branch) and came to an end about 37 miles from the base of the range at an altitude of 2,500 feet. The low levels reached by these trunk glaciers seem truly remarkable when it is considered that their lower parts lay wholly in the zone of wastage where, even in the shaded spots, summer heat was suflicient to remove the snows of winter. The Yosemite glacier, for in- stance, reached more than a mile below the level (some- what above 8,000 ft) in which glaciers were formed in the Yosemite region. The Tuolumne glacier reached 6,000 feet below this level; the San Joaquin glacier, about 5,300 feet; the Kings glacier, about 6,000 feet. (The level of glacier generation rose gradually southward.) The ability of these glaciers to reach such low levels in spite of the warmth that prevailed in the zone of wastage affords impressive testimony of the immense surplus of snow and ice that descended from the higher parts of the range. However, these low levels were also due in part to the protection from the sun’s rays that was afforded to the glaciers by the high walls of the canyons; to the small surface areas, proportionate to bulk, that the glaciers, 3,000 to more than 4,000 feet in thickness, presented to the melting agencies; and to the relatively rapid movement of the ice, which in the thicker glaciers must have averaged several feet a day. On its eastern flank also, the Sierra Nevada bore a great array of glaciers, there being a glacier in almost every canyon; but these glaciers were in general much shorter than those on the western flank, owing to the abruptness of the escarpment, the shortness of the can- yons, and the small extent of glacier-generating terri- tory at their heads. Most of these glaciers, neverthe- less, reached down to the eastern foot of the range; not a few projected well out into the adjoining lowlands, especially in the north-central and south—central parts of the range, where these lowlands have altitudes of 5,000 to nearly 7,000 feet. The basin of Mono Lake was invaded by no less than six ice tongues, each of which extended several miles out from the range. In the regions south of Mono Lake, the glaciers projected as a rule but little beyond the mouths of their canyons, and along the border of Owens Valley the glaciers were confined mostly to the upper A19 parts of the canyons; still further south there were only scattered snowfields, and the array of ice bodies came to an end. BASIS FOR DIFFERENTIATION OF THE GLACIAL STAGES The courses of the ancient glaciers in Sequoia Na- tional Park were traced and mapped and their farthest limits were determined by the same method that had proved effective in the San Joaquin Basin and else- where, that is, interpretation of the testimony of glacial deposits rather than that of sculptural features; this method consists primarily of a systematic survey of the moraines that were built by the individual glaciers. In open country such a survey can readily be executed with sufficient accuracy for a. reconnaissance map by locating the moraines by eye with respect to identifiable landmarks, of which the landscape of the Sierra Nevada affords a plenty; but in the forested tracts the larger moraines must be actually followed out and, in some places, located by traverse—a laborious and time—con: suming process. Fortunately, in Sequoia National Park the forested areas, though of considerable extent, are so amply diversified by topographic and drainage features, as well as by occasional meadows, that a large share of the work could be done by following the moraines, or the swales between moraines, on horseback. In the rougher areas, of course, the mapping had to be done on foot. In the morainal deposits of Sequoia National Park, as in the areas to the north previously mentioned, abundant and unmistakable evidence was found of two distinct stages of glaciation-a later one, the Wisconsin, and an earlier one, El Portal, separated by a lengthy time interval; meager indications were also found of a third, very early stage, the Glacier Point (Matthes, 1929). Thus the observations on the morainal deposits of these different stages in Sequoia National Park bear out the correctness of the analysis that was made of the moraine systems of the other basins in the Sierra Nevada, and place on a firmer basis the author’s inter— pretation of the succession of the events in the glacial history of the Sierra Nevada. Throughout the central and south-central Sierra Nevada, therefore, the moraine systems of the great trunk glaciers and their numerous branches spell out the same story of three distinct pe- riods of extensive and long-continued glaciation dur- ing the Pleistocene Epoch. The characteristics of the glacial deposits of the three stages found in Sequoia National Park and else- where on the west slope of the Sierra Nevada are set forth in the following sections. A20 WISCONSIN STAGE The glaciers of the Wisconsin Stage are considered first, because they are most definitely known and were, for the most part, ice tongues confined to individual valleys. Having obtained a definite image of them, the reader can then visualize also the more extensive glaciers and ice fields of the earlier stages, which coalesced over divides and in part moved in disregard of them. Throughout the Sierra Nevada the moraines of the Wisconsin Stage are, as a rule, well preserved and dis- tinct. Many still retain, only slightly changed, the sharp-crested forms which the glaciers gave them (fig. 20). Much of the finer material has been washed from these moraines, but many boulders that form the crests remain in place, or substantially so. The frontal mo— raines are commonly breached by the streams, but some of the younger moraines still act as dams impounding lakes. On slopes of low or moderate declivity the lateral moraines are often splendidly developed, ex~ ! x. . ,1 3 . ,4”! It .. SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY tending for long distances as regular embankments broken by only a few stream-cut notches. Though wholly absent on precipitous canyon walls, lateral mo- raines can usually still be traced along steep slopes by surviving patches of coarse debris or by single boulders. Particularly useful as diagnostic materials are the boulders of different types of granitic rocks. Such rocks preponderate in the central and southern parts of the Sierra Nevada and make up the bulk of the moraines (figs. 21, 22). In the moraines of “7isconsin age, a large percentage of these granitic boulders are unweathered and unstained. They ring when struck with the hammer, and with fine resilience throw the hammer sharply back; this is true not only of the sili- ceous types of granite and granodiorite but also of the more basic rocks—the quartz diorite, diorite, and gab- bro. Exception must be made, of course, for such boulders as were already weathered when picked up by the glacier. FIGURE 20.—One of the timbered moraines of the Wisconsin Stage that surround Moraine Lake. Note horse and rider for gaging size. The meadow, which is below Moraine Lake, occupies a strip of level swampy land formed by the gradual filling of a lakelet that lay between the moraine in view and Granite sand continues to be the next one to the right. washed down from the flanking moraines, but the meadow is still too wet for the growth of lodgepole pines. A few seed- lings are beginning to invade it. GLACIAL A21 FIGURE 21,—A frost-split block of granite on one of the Wiscon- sin moraines that encircle Moraine Lake. Measurement of the pieces shows that originally the block was 23 feet long. The glacial deposits of Wisconsin age are associated with smoothed and polished floors, walls, roches mou- tonnées, and ledges of rock in place (figs. 23, 24, 25). These surfaces are most extentive in the areas of sparsely jointed siliceous granite, and surfaces that were glaciated during the later substages of the Wiscon- sin naturally are more perfectly preserved than those that were glaciated during the earlier substages. From the latter surfaces, usually, most of the polish has al— ready flaked ofi', but the extent to which it has disap- peared depends also in some measure on the character of the rocks, the acid types retaining the polish longer than the basic ones. Whether still polished or not, how— ever, all rock surfaces planed down by the Wisconsin ice still exhibit today the smoothed forms that are well known to be characteristic products of glacial abrasion. The actual reduction effected by weathering ranges from virtually none to as much as 2 inches. The signs of recency in the Wisconsin moraines, whether preservation of ridge forms or freshness of boulders, are likewise somewhat more accentuated in the later than in the earlier Wisconsin deposits. In the somewhat more subdued moraines of the earlier substages, although the acidic rocks are still generally sound and hard, an increasing percentage of the basic It has fallen apart as a result of the force exerted by water freezing in incipient joints. rocks are cracked or split along joint planes, and some of these rocks are so badly disintegrated that they crum- ble in the hand. However, there is no need in this study to difi'erentiate between substages of the Wiscon— sin Stage. The Tahoe and Tioga glaciations, which have been distinguished on the east slope of the Sierra Nevada and regarded as stages of early and late Wiscon— sin age, respectively (Blackwelder, 1931), are now real— ized to be two distinct stages in the Sierra Nevada. Correlation of the last stage of Pleistocene glaciation in the Sierra Nevada with the Wisconsin Stage of the Laurentide glaciation was based initially on the compar- able degree of preservation of their respective morainal deposits and on the largely unweathered state of the boulders in them. But there is additional warrant for the correlation in the fact reported by Alden (1932) that in northern Montana, where the Keewatin drift of Wisconsin age overlaps the moraines of the last Cordil- leran glaciation, both deposits have about the same de- gree of freshness. There can be no doubt, on the strength of this evidence, that the last Pleistocene glaci- ation of the Rockies and the last Pleistocene glaciation of the adjoining plains were contemporaneous. It is assumed, further, that the last Pleistocene glaciation of the Sierra Nevada occurred at about the same time as A22 SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY FIGURE 22.——Disrupted glacial boulder in Wallace Canyon. of the Wisconsin Stage. The boulder, when intact, measured 8 feet wide and 5 feet high and doubtless was in one piece when deposited by the glacier toward the end Opening of the originally tight joints in the granite by the freezing of infiltrated water has resolved the boulder into a series of parallel slabs. that of the Rockies; though incontrovertible proof of this synchronism is lacking, there is at present no known reason for thinking otherwise. EL PORTAL STAGE The ill—defined round-backed moraines of El Portal Stage stand in striking contrast with the distinct sharp- crested moraines of the Wisconsin Stage in the Sierra Nevada. Disintegration of the boulders in El Portal moraines has continued for so long a time as to destroy completely the original crests. It has resulted also in mantling these older moraines with considerable arkose sand that gives them a smoother and less bouldery ap— pearance than the Wisconsin moraines usually have. Although far more bulky, El Portal moraines are rela- tively obscure features that may easily be overlooked by an untrained observer of the landscape. To identify these obscure moraines beyond possible doubt, their constituent materials must usually be sought in gullies, roadcuts, trailcuts, or holes left by uprooted trees. Although gullies are ordinarily plentiful, the author was obliged more than once to rely wholly on the evidence furnished by uprooted trees. Fortunately, some of these trees held boulders and cobbles aloft in their exposed roots, as if for the convenience of the geologist. Almost invariably such boulders and cobbles are light buff or yellowish because of their limonite coat- ting. The entire deposit of which they form part is commonly of the same yellowish hue, in contrast to the morainal material of Wisconsin age, which is mostly gray. The coating on the boulders and cobbles general- ly masks their lithologic character and makes many of the granitic rocks indistinguishable. When broken open, the boulders and cobbles are usually found to be stained by ferric oxides to a depth of one—fourth to one-half inch. The boulders composed of the more siliceous granites are as a rule still firm, but they have lost so much of their resilience that the hammer bounces back only feebly from them. Many break readily; some even crumble into discrete granules. The granodiorite and quartz diorite boulders usually have still less coherence, and the diorite and gabbro rocks are so weak that they are smashed to bits by a light blow. These boulders are often traversed by ramifying cracks; some are in a crumbling stage. These characteristics vary, of course, with the prevailing moisture conditions in the deposits. In well-drained locations, decomposition and disinte- gration make much slower progress than in poorly drained ones. On strongly isolated platforms of bare GLACIAL RECONNAISSANCE OF SEQUOIA NATIONAL PARK A23 FIGURE 23.~—-Glacier polish, striae, and grooves, above the head of Kern Canyon. The rock is aplite, which weathers more slowly than the coarser granite and therefore holds its glacial rock, which dry out quickly, scattered cobbles and boul- ders are apt to be surprisingly well preserved (figs. 26, 27). Boulders on top of the moraines, exposed to the heat of the sun and therefore drying quickly after a wetting, generally have some cohesive strength left, but those in the interior of the moraines are as a rule so weak that the picks of the road workers cut right through them. In some roadcuts such boulders, cut flush with the wall, appear outlined as rusty rings. Terminal moraines of El Portal Stage are generally wanting in the main canyons of the west slope of the Sierra Nevada. Their absence can hardly be attributed to complete destruction by stream-and-weather erosion after El Portal time, for in some of the canyons the topography is decidedly favorable to the preservation of at least the wings of such moraine. A more probable explanation is that the trunk glaciers of El Portal Stage built either no terminal moraines at all or else only very small ones, because the ice fronts rested for no considerable length of time at any point during the maximal phases. Whatever the explanation, the far- thest limits reached by the trunk glaciers of El Portal 748—960 0—65—45 Since being glaciated, the aplite has been markings longer. somewhat disrupted into angular blocks by repeated frost action. Stage cannot, as a rule, be determined with any great accuracy because of the lack of terminal moraines. The lateral moraines left by the glaciers of El Portal Stage, on the other hand, are generally of massive pro- proportions that dwarf the corresponding laterals of the Wisconsin Stage. Though gashed or even transected by gullies, and though in some places almost destroyed, they are nevertheless not diflicult to trace, once they have been identified. It is, indeed, by the systematic mapping of these old lateral moraines that the courses of the glaciers of El Portal time are most surely traced. It may appropriately be added that for polished and striated rock surfaces dating from El Portal Stage one need not search. The glistening floors, walls, ledges, and roches moutonnées that are so abundant and so impressive in some parts of the Sierra Nevada, as al- ready noted, all date from the Wisconsin Stage— mostly from its later substages. The rocks surfaces that were planed by El Portal ice have long since been destroyed and changed beyond recognition by the gran- ular disintegration, scaling, or spalling of the rocks A24 SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY FIGURE 24.—Fantastic rock forms in the upper Kern Basin, near Lake South America. The forms were produced by glacial quarrying of joint blocks followed by abrasion and rounding of the resultant angular forms. The direction of ice move- (figs. 28, 29). In many places the granitic rocks ap— pear to have been stripped of disintegration products to depths of 10 to 12 feet. Approximate measurements of such stripping are afforded by residual rock pedes- tals supporting perched glacial boulders and by dikes of slow-weathering aplite that remain standing in high relief (Matthes, 1930, p. 70—74; 1950a, fig. 51; 1950b, p. 101—102; 1960, fig. 32). Several other types of cri— teria suggest themselves, notably the depth of stream channeling accomplished since El Portal and Wisconsin climaxes, respectively, and the erosional changes pro- duced in valleys and other landforms; but these criteria generally furnish no more definite measures for com- parison than the pedestals and dikes. This marked diflerence in the depth of disintegration and stripping of the granite since El Portal Stage and since the Wisconsin Stage aflords a good index of the relative antiquity of those two stages of glaciation. If the time since the climax of the Wisconsin Stage is to be measured in tens of thousands of years, then surely the time since the climax of El Portal Stage is ment was diagonally toward the right and away from the camera. Many blocks that were firmly attached when the glacier passed over them have since been split or loosened by postglacial frost action. to be reckoned in hundreds of thousands of years. Mainly for that reason the author holds that El Portal and its probable correlative on the east side of the Sierra Nevada, the Sherwin Stage (Blackwater, 1931), are not younger than the Illinoian Stage of the conti- nental glaciation and may include deposits of two or more unseparated stages, perhaps both the Illinoian and Kansan (Matthes, 1933, p. 33). GLACIER POINT STAGE The earliest glaciation that has been recognized on the west slope of the Sierra Nevada, the Glacier Point Stage, which presumably corresponds to Blackwelder’s McGee Stage on the east flank (Blackwelder, 1931), is indicated by morainal deposits in only a few localities in the range. Two circumstances account for the scar- city of these deposits: (1) Their obliteration over large areas by the extensive and voluminous deposits of El Portal Stage, and (2) their vestigial character, their constituent materials having in large part disintegrated and been carried away. GLACIAL RECONNAISSANCE OF SEQUOIA NATIONAL PARK A25 FIGURE 25.———View across the glaciated floor of Whitney Canyon, showing combined effects of quarrying and grinding. The glacier moved from left to right, approximately parallel to a set of vertical joints in the granite. Horizontal joints enabled The few small deposits which the author has shown on his maps as probably belonging to the Glacier Point Stage lie, in most places, at levels 200 to 300 feet above the upper limit of the massive El Portal moraines, where they clearly have escaped obliteration. Their positions at those high levels, however, do not necessarily indicate that the ice streams of the Glacier Point Stage attained greater depth in the canyons than did the ice streams of El Portal Stage; for it stands to reason that during Glacier Point time the canyons were not yet cut to the depth which they later attained in El Portal time. A given quantity of ice would have filled them to a higher level in Glacier Point time than it would have in El Portal time. Some of the deposits of the Glacier Point Stage lie far out on the uplands flanking the canyons, and thus show that the ice of that stage overflowed the canyon rims, whereas the moraines of El Portal Stage show that the glaciers of that stage remained largely confined within the canyons. But that fact, too, is very probably explained by the circumstance that in Glacier Point time the glacier to quarry out long slabs, but this quarrying pro ceeded slowly because of the scarcity of vertical cross joints. As a consequence, the glacier could grind and round off many of the slabs before tearing them out. the canyons had smaller cross—sectional areas, and there— fore less capacity for holding ice, than they had later in El Portal time. In contrast to the massive El Portal moraines, the deposits of the Glacier Point Stage have but meager volume and a decidedly depleted aspect. They are re- duced for the most part to skeletonlike rows of erratic boulders composed of resistant quartzite and highly siliceous granite, the rest of the constituent materials having vanished. In some places the deposits are en— tirely destroyed. The weaker rocks, such as the dio- rites, are represented as a rule only by a chance frag— ment here or there. Many of the siliceous boulders, even, have lost their glacial contours as the result of spalling, exfoliation, or granular disintegration; some have been reduced to strangely cavernous or basined forms. The fact that the morainal materials have been transplanted can be established only on lithologic grounds and by a knowledge of their provenance. Need- less to add, it requires a trained eye to identify such scanty, vestigial deposits; and to know where to look A26 FIGURE 26.——-A 16-foot erratic left on Bighorn Plateau by the ice of El Portal Stage. No continuous moraine exists here, only scattered ice-borne boulders. Many boulders are in process of breaking up; others have already disintegrated into granite sand. The fragments at the base of the large boulders are spalls split from it by frost action. for them, one must have some conception of the extent of the earlier ice in the Sierra Nevada and also a knowl- edge of the habits of glaciers and the manner in which they adjust their movements to different types of topography. SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY That the moraines of the Glacier Point Stage are much older than the moraines of El Portal Stage is readily evident from the foregoing. Even if it be. granted that the relative scantiness of the Glacier PO! 1t deposits may be due to the shorter duration of the Glacier Point Stage as compared with El Portal, it is manifest from the more advanced state of disintegration of the boulders in the Glacier Point moraines that the Glacier Point Stage preceded El Portal Stage by a considerable length of time. The Glacier Point Stage clearly is to be assigned to the early Pleistocene, and some argument may even be found for correlation with the Nebraskan, for the Glacier Point is the earliest stage of glaciation of which any recognizable deposits remain in the Sierra Nevada (Matthes, 1933, p. 33). GLACIATION OF THE KAWEAH BASIN: KAWEAH GLACIER SYSTEM In both the Wisconsin and El Portal Stages of the Pleistocene, the numerous converging canyons of the Kaweah Basin became pathways for cascading ice streams. However, inasmuch as even the larger of these streams attained lengths of only 10 miles, they fell short by many miles of uniting into a single Kaweah Basin glacier system or of forming a major trunk glacier com- parable to the ones in the main drainage basins to the 1 FIGURE 27.——Glacial boulder of El Pertal Stage resting on a platform overlooking the Big Arroyo. Weathering has produced a bread-crust elfect on the sides and weather pits 6 to 15 inches deep on the top surface. GLACIAL RECONNAISSANCE or SEQUOIA NATIONAL PARK A27 FIGURE 28.—Glaciated knob at the head of South Fork, Kaweah River. This knob was overridden by the earlier glaciers but not by those of the Wisconsin Stage, as is evident from the relative position of the older and younger moraines nearby. During the long period since it was glaciated, the knob weath- FIGURE 29.—Saddle east of Tower Rock, on the east rim of Kern Canyon. This saddle was invaded by the Kern glacier of El Portal Stage. The crags and boulder in this View, which ered into jagged forms. Infiltration, of water doubtless has been facilitated by the high angle of the jointed fractures, and, as a consequence, disruption by frost has been particularly vigorous. are 10 to 20 feet high, give evidence of the post-El Portal weathering that here has destroyed all traces of glaciation. A28 north and in the Kern Basin to the east. Because only the ice streams of adjacent tributary canyons united, there formed, instead, a series of separate glacier sys- tems of relatively small extent—one or more in the headward areas of each of the main branches of the Kaweah Basin: the Marble Fork, the Middle Fork, the East Fork, and the South Fork. These were branching glaciers that only locally, in their upper reaches, over— swept the divides; elsewhere they were confined to the canyons, which were filled only in part. The lower limits of the Wisconsin glaciers are clearly marked, for the most part, by distinct moraines, and these indicate that the termini of the principal ice streams reached altitudes as low as 5,200 to 8,200 feet. The exact extent of El Portal glaciers is unknown and may never be known, for terminal moraines are lack- ing, and the lateral moraines are partly concealed by dense Chaparral on the lower slopes of the canyons. However, the approximate lower limits of glaciation can be inferred from the available evidence, and, for the principal ice streams, they lay at altitudes of 4,550 to 6,500 feet—marking limits which were considerably lower, therefore, than the corresponding ones of the Wisconsin glaciation. The limited development of the Pleistocene glaciers of the Kaweah Basin is not surprising, since this basin heads not along the main Sierra crest, as do the adjacent Kings and Kern Basins, but rather along secondary ridges 12 or more miles farther to the west—that is, this basin extends only part way up the Sierra west slope. To be sure, at the east edge of the Kaweah Basin the glacial sources lay along the bold Great Western Divide, whose peaks have altitudes of 11,000 to over 12,500 feet, not greatly inferior to those of the main crest. But elsewhere the glacial sources of the Kaweah Basin lay along secondary ridges which branch off from the Great Western Divide and are, for the most part, much lower. One of these ridges, which includes a segment called Silliman crest, extends northwestward from Triple Divide Peak and forms a part of the north- east rim of the Kaweah Basin. From this ridge, at altitudes of 10,000 to 11,600 feet, several ice streams descended into the northeasterly part of the Kaweah Basin. Another ridge, winding southwestward and southward from Florence Peak, forms the east rim of the South Fork Basin. Though for the most part only 9,000 to 11,500 feet in altitude, this ridge also gave rise to glaciers, and these descended into the canyons of the East Fork and the South Fork. Only in the lower Kaweah Canyon were deposits found which appear to represent the earliest, or Glacier Point, stage of glaciation. Location of these deposits is not shown on plate 1 but is indicated on page A30. SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY The deposits, being outwash materials laid down a con- siderable distance beyond the termini of the glaciers, give little beyond a suggestion concerning the nature and extent of this ancient glaciation in the Kaweah Basin. MARBLE FORK WISCONSIN STAGE The Marble Fork glacier was one of the principal ice masses in the Kaweah Basin, reaching a length of over 9 miles and extending down Kaweah Canyon to an altitude of about 5,350 feet. The trunk glacier, in the canyon, was fed by several short tributaries from the south, which originated at altitudes of 10,000 to more than 11,000 feet on the upland bearing Alta Peak (fig. 30), and by several considerably longer tributaries from the northeast, which originated on the ridge which bears the Tableland (11,000—11,600 ft) and Silliman Crest (10,000—11,200 ft). Because of the disparity in size between the southern and northern tributaries, the glacier system as a whole had an asymmetrical pattern. The northern tributaries were confluent, across a few low places on the divide, with the heads of the Kings River glacier system; and southeast of Table Meadows an icefield 21/2 miles wide connected the Marble Fork glacier with the Buck Canyon glacier, a member of the Middle Kaweah glacier system. The valleys of Silliman Creek and Clover Creek also contained glaciers, but because these were shallow ice bodies only 21/2 and 3 miles long, respectively, they did not join the trunk glacier. Wisconsin moraine plasters the slopes along the road connecting Giant Forest and Lodge Pole and across the valley from this road. In upper Silliman Creek valley, massive embankments of Wisconsin moraine lie on both sides of the creek, and a high steep-fronted moraine that flanks the lower edge of Cahoon Meadow is trenched by the creek. In Clover Creek Valley also, the main loop of Wisconsin moraine is complete, extend- ing back as far as the mouth of the ungl‘aciated West Fork. The effects of vigorous glacial erosion in predomi- nantly massive exfoliating granite are strikingly shown both in the clean-swept area extending from Tokopah Valley to the Tableland and in the vicinity of pictur- esque Heather, Pear, and Emerald Lakes, which lie in compound cirques. These lakes are typical tarns and are held in by barriers of massive granite. In places on the cirque walls back of these lakes the upper limit of glaciation is very plain; sheer cliffs of frost-shattered granite come down to the smoothly sloping platform that was abraded by the moving ice. The bowl of the cirque is very imperfect because of the many ledges of massive granite that could not be eroded away. The GLACIAL RECONNAISSANCE 0F SEQUOIA NATIONAL PARK A29 FIGURE 30.—A1ta Peak (11,211 ft). The summit is composed of frost-shattered remnants of exfoliation shells. Formerly the shells extended toward the left in a descending curve out- lining a dome, but the excavation of a cirque by a small rOcks have been rounded and smoothed to pillowlike forms. Aster Lake, the small tarn below Emerald Lake, provides a good example of selective quarrying, being itself encased between masses of solid, very sparsely jointed granite. The little valley southwest of Heather Lake held a small glacier that did not quite join the master glacier in Tokopah Valley. This valley heads in a poorly shaped cirque which may be described as a veritable glacial quarry. There the glacier left enormous quanti- ‘ties of quarried blocks in a terrific jumble, gleaming white among the forest trees. . EL PORTAL STAGE ‘ The Marble Fork glacier in El Portal Stage was over 10 miles long and 25 miles square, and it reached down Kaweah Canyon to an altitude of 5,350 feet. It was joined, far downstream, by its major afliuents, the Silli- man Creek and Clover Creek glaciers. That these gla- ciers were confluent with the trunk glacier in El Portal Stage is indicated by remnants of older drift scattered . oyer the lower slopes of both valleys. glacier has pared away the north side of the dome, thereby giving the summit the unsymmetrical profile seen in this view. The upper part of the cirque wall, which is several hundred feet high, is visible in the lower left corner. The road from Giant Forest to Lodge Pole crosses the fairly heavy and bouldery left lateral El Portal moraine near the Wolverton Creek bridge; on the north side of the valley, in the vicinity of Willow Meadow, the J. 0. Pass Trail crosses the right lateral moraine. In- asmuch as Willow Meadow lies among older moraines, some of them fairly prominent, it is evident that this divide was overswept by the earlier ice. Seven miles below the terminus of the Marble Canyon glacier—that is, about half a mile above Ash Mountain headquarters—a deposit of material interpreted as bouldery outwash of El Portal Stage is revealed in a roadcut on the Generals Highway (Mat-thes, 1950a, p. 52).2 (See fig. 31.) This material is believed to have been washed down from El Portal glaciers in the upper canyons, probably not only those in the Marble Fork but also those in the Middle Fork. The material was deposited in the streambed but, as a result of continued trenching by the river, it is now about 100 feet up the canyon side. Evidently this deposit is of considerable aThe above is apparently Matthes’ only reference to this deposit. No mention of it appears in his field notes. F. F. . A30 , SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY FIGURE 31.———Outwash of El Portal Stage, revealed in a roadcut 0n the Generals Highway, about half a mile above Ash Mountain Headquarters. age, for all the boulders have rusty surfaces, some of them are partly decomposed, and the interstitial sand is stained reddish brown by iron oxide. GLACIER POINT STAGE3 In the lower Kaweah Canyon, considerable masses of material, presumably outwash of the Glacier Point Stage, remain. (See fig. 32.) These masses occur as inconspicuous terraces on the side of the canyon about 200 feet above the river bed, and they are exposed in several roadcuts along the Generals Highway. The great antiquity of these deposits is evident from the fact that all the boulders except the surface ones are completely decomposed and can be cut through like so much granite sand. In making these roadcuts, no blast- ing or crowbar work was necessary; a Civilian Con- servation Corp crew with picks and spades cut with 3 Information in this section is taken from Matthes (1950a, p. 53) and from a letter that Matthes wrote to Robert W. Sayles dated Novem- ber 29, 1940. The letter in one place refers to the deposits as “till,” but this was evidently a slip. for they are otherwise termed “outwash.” Location of these deposits is not shown on Matthes’ maps, nor is there reference to them in his field notes. 1F. F. ease through the decayed boulders and the interstitial sand and shaved them back to a uniform, smooth slope. Rainwater rills have since carved little furrows in tha slope, trenching both the boulders and the matrix aroum them to equal depth. Paradoxical as it may seem, thu‘ tops of the uppermost boulders, which project slightly above the surface of the deposit, still remain firm ang would require a blow with a sledge hammer to be broken. They survive in the form of convex caps because thejil are in well-drained positions and are frequently ex posed to the drying rays of the sun. MIDDLE FORK i During both the earlier and the later glacial stages, the converging valleys at the head of the Middle Fork Canyon were occupied by ice streams that attained sufficient length to unite into a short trunk glacier it the main canyon, below Redwood Meadow. (See fig‘t 33.) The glaciers of both stages were confined to their respective canyons, nowhere overflowing onto thjt higher parts of the intervening divides. The larges \ “mow GLACIAL RECONNAISSANCE OF SEQUOIA NATIONAL PARK A31 FIGURE 32.——G1acia1 outwash, dating perhaps from the Glacier and now forms an inconspicuous terrace on the side of the Point Stage, exposed in a roadcut on the Generals High— canyon, about 200 feet above the riverbed. Photograph way above Camp Potwisha. This material was brought by J. C. Patten. down by the Kaweah River from the glacier at its head FIGURE BEL—View up “River Valley,” the glaciated upper canyon of Middle Fork, Kaweah River. Clilfs in the foreground show exfoliation. A32 ice streams, those of the Middle Fork proper and of Clifl' Creek, attained lengths of approximately 10 miles in the earlier stage and 9 miles in the later stage. The principal sources of ice lay to the east, in cirques on the Great Western Divide; however, several other sources lay on the secondary divides to the north and south. In the earlier stage, the lowest altitude reached by the trunk glacier, in Middle Fork Canyon, was about 4,800 feet; in the later stage, about 5,200 feet. Moraines deposited by the earlier Middle Fork glacier system are found in the vicinity of Redwood Meadow, which is in the triangular area between Middle Fork Creek and Cliff Fork Creek. The grove of Big Trees which. gives the name to this meadow stands mainly on older drift, and old moraines also enclose the grove on both sides. A second grove, about a mile farther up Clifl' Creek on the north side of the stream, stands partly on older moraine and partly above its limits. Wisconsin moraines are also conspicuous in the vicin- ity of Redwood Meadow. The trail from Redwood Meadow to Little Bearpaw Meadow crosses the left lateral moraine of Middle Fork glacier. Farther to the southeast is the right lateral moraine of the Cliff Creek glacier. Big Trees stand on both of these ridges. In Middle Fork valley, which is littered with Wis- consin morainal debris, the trail from Little Bearpaw Meadow descends through the depression back of the right lateral moraine, a ridge about 100 feet high. The SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY trail crosses the moraine by way of a gap cut by the stream flowing from the meadow. Southwest of the gap, the moraine is less prominent. On the opposite side of the valley, the Wisconsin moraines have a dis- tinct upper limit, above which are smooth mountain slopes densely covered with brush. In this section of the valley, glacial outwash forms terraces about 50 feet high on both sides of the Middle Fork. The Hamilton Lakes locality, in the upper reaches of the Middle Fork Basin, impressively illustrates the erosional effects of glaciation and concomitant snow avalanching (fig. 34, frontispiece). The large lower lake (altitude 8,300 ft)—occupying a rock bowl with a smooth glaciated lip, clearly visible under the water— marks the place of confluence of glacial tributaries that issued from three cirques lying 1,700 to 2,000 feet higher. These tributaries excavated the lake basin in what is really a canyon step. Northwest of the lake, massive granite, cleft by nearly vertical joints, has been sculptured into pinnacles somewhat El Capitan-like in aspect, but more complex (fig. 35). In the vicinity of this lake, the effects of snow avalanches are evident on every hand. One avalanche chute northeast of the lake outlet has not only an avalanche cone beneath it but also a dump in the lake, visible from the trail above. In the valley of Cliff Creek, terraces of glacial out- wash 50 to 60 feet high occur near the junction with Timber Gap Creek. The trail enters the valley of FIGURE 34.——Spectacular summits south of Hamilton Lakes. They have been produced by glacial sculpturing in massive exfoliating granite. z GLACIAL RECONNAISSANCE 0F SEQUOIA NATIONAL PARK FIGURE 35.-—View from near the source of Hamilton Creek, down the canyon across one of the Hamilton Lakes. Massive granite forms the impressive cliffs at the right and the rock Timber Gap Creek by climbing steeply over a gigantic moraine. Through this ridge the creek has cut a nar- row gulch with steep bouldery sides. In addition to the main glacier system, independent ice bodies of small size developed in the Middle Fork Basin during both glacial stages. On the north side of the basin, glaciers formed in upper Mehrten Creek Valley, Buck Canyon, and the two intervening valleys. On the south side of the basin, a single small glacier formed in the valley of Little Sand Meadow. At the head of Mehrten Creek, southwest of Alta Peak, Wisconsin glaciation is recorded by an imperfect cirque containing a small quantity of jumbled morainal material. On the slopes below this cirque, along the Sevenmile Hill trail, are scattered morainal materials of the earlier glaciation. In the adjacent valley, which contains Alta Meadow, the glacial record is much clearer. Here, at the lower margin of another shallow cirque, lie at least three Wisconsin moraines. They are composed of angular rock blocks, and each morainal loop protects a sloping shelf on its upslope side. Below these moraines, de- barrier across which the lake has its outlet. In the center of the picture on the distant mountain, is a well-formed avalanche chute. Photograph by W. L. Huber. posits of the earlier stage extend far down the moun- tainside. The High Sierra Trail crosses the older moraines, which are cut by a steep—sided gulch. Alta Peak, the summit which bore these glaciers, has a highly asymmetric profile, its northern slopes having been vigorously glaciated and its southern slopes only mildly so (fig. 30). A considerably larger glacier occupied Buck Canyon during both glaciatiOns. As previously noted, the broad icefield at the head of this glacier was confluent with the Marble Fork glacier system. In the earlier stage, the ice descended the canyon a distance of 5 miles, to an altitude of 5,600 feet. The tapering ice tongue of this glacier was joined by a tributary, more than 3 miles long, that originated in a cirque northeast of Alta Peak. In the later stage, the Buck Canyon glacier was some- what shorter, and its terminus reached only to 6,000 feet. The tributary from northwest of Alta Peak was then only 11/2 miles long and therefore remained a separate glacier. The High Sierra Trail, from Giant Forest to Bearpaw Meadow, crosses Buck Creek and ascends the Wisconsin left lateral moraine deposited by the Buck A34 Canyon glacier—a fairly sharp moraine—and then traverses older moraines plastered against the valley- side. The latter moraines, however, do not reach to the top of the ridge between Buck Canyon and the valley of the Middle Fork. EAST FORK Glaciation occurred during both the Wisconsin and the El Portal Stages of the Pleistocene in the upper basin of the East Fork, not only in the main valley it- self (in the Mineral King region, outside of Sequoia National Park) but also in the large tributary valley of Horse Creek. The main valley held a many-branched glacier which, in the earlier stage, was almost 10 miles long and de- scended to 4,950 feet and, in the later stage, was 6 miles long and descended to 6,850 feet. The trunk glacier was fed by long, narrow tributaries, some originating at the east, in cirques on the Great Western Divide, but most originating at the south, on the secondary ridge extend- ing westward from Florence Peak on the Great West- ern Divide. In the asymmetary of its pattern, the gla- cier system of the East Fork somewhat resembled that of the Marble Fork, with the obvious difference that the principal sources of the East Fork were on the south side of the basin instead of on the north. Indeed, the amount of ice contributed to the East Fork glacier sys- tem from the north was negligible. During the Wisconsin Stage the most westerly tribu- taries of the East Fork glacier system remained as small independent glaciers. Two of these were in the upper valleys of Deer Creek. Older moraines of the East Fork glacier occur along the road which ascends the north side of the valley. They begin at a point about 1 mile east of Silver City and continue eastward for a distance of more than 2 miles. The road enters Wisconsin deposits near the junction of the Mosquito Lakes outlet with the East Fork, and bouldery moraines of this age continue up- stream to the bend of the river near Mineral King. In places these moraines have been trenched by the river. The absence of conspicuous moraines in some of the up- per tributary valleys, as in the vicinity of Farewell Gap, may be attributed to two circumstances: the small size of the rock fragments derived from the metamorphic rocks which prevail here, and the changes wrought by periodic snowslides. The trail to Eagle Lake climbs a fairly smooth slope of moraine consisting mostly of small fragments and slabs of metamorphic rocks and only a scattering of granitic boulders. This entire val- ley appears to be sheathed with such materials. The earlier Horse Creek glacier was about 61/2 miles long and descended to 5,700 feet; the later‘one was about 5 miles long and descended to 6,650 feet. Most of the SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY ice came from broad, shallow cirques to the northeast and east, but in the earlier stage, ice from the valley of Whitman Creek spilled over the north edge of the Hockett Meadows northwest of the Hockett Ranger Station, cascaded down the slope several thousand feet, and, 5 miles from its source, joined the Horse Creek glacier. In the Wisconsin Stage the Whitman Creek glacier fell just short of uniting with the Horse Creek glacier, as indicated by the moraines which extend to the lip of the Whitman Creek valley and describe their farthermost loop where this valley begins to drop off from the level of the plateau. Northeast of Evelyn Lake is a little cirque in the lee of a summit only 8,900 feet in altitude (the actual rock rim of the cirque is at 8,700 ft.). In the earlier stage, this cirque contributed ice to the Horse Creek glacier; in the later stage, the cirque held a small independent ice body. The lower trail to Cahoon Meadow crosses the Wisconsin moraines of this glacier. Evelyn Lake occupies a steep-walled cirque which held, during both stages, a glacier formed in the lee of a plateau summit 9,100 feet in altitude. Large Wisconsin moraines com- posed of big angular blocks border the lake and extend about half a mile down the canyon; below these mo- raines lie older ones which are crossed by the trail to Cahoon Meadows. The features at Evelyn Lake well illustrate the effectiveness of erosion by even a small glacier. Furthermore, this glacier and its small neigh- bor are of special interest because of the low altitudes at which they originated. SOUTH FORK The upper South Fork Basin was occupied, during both glacial stages, by an irregularly shaped compound ice mass. In the earlier stage this glacier not only dis- charged down the canyon of the South Fork but also was continuous with the Whitman Creek glacier north- ward through Hockett Meadows. The South Fork was then 81/; miles long, measured from its source above the Blossom Lakes to its terminus at 6,200 feet in the South Fork Canyon. In the later stage this glacier was a mile shorter and extended down only to 7,500 feet. This broad, shallow ice mass left a covering of mo— raines and scattered debris on the country across which it spread. Some of these deposits are crossed by the trail from Wet Meadow to Hockett Ranger Station. Ice advancing from the vicinity of Sand Meadow scooped out the shallow basins of the Hockett Lakes, overrode the ridge to the west of these lakes, and spilled down into South Fork Canyon. From the evidence of glacial boulders which strew Tuohy Meadow and the South Fork Meadows, it is clear that all of this part of the South Fork basin was covered GLACIAL RECONNAISSANCE OF SEQUOIA NATIONAL PARK by ice streams converging from the numerous pockets in the high southern and southeastern rims of the plateau. During both stages, a small isolated glacier lay on the north side of the peak (9,405 ft.) at the northeast end of Dennison Ridge. The trail leading from the Hockett Lakes to the Garfield Grove of Big Trees crosses the moraines of this glacier. The so-called “sand dune” which gives name to Sand Meadows is a nearly circular, rounded deposit ceihposed of sand, subangular gravel, cobbles, and boulders as much as 9 inches in diameter (fig. 36). The dune is ob- viously not aeolian but is the remnant of an outwash de- posit formed late in the Wisconsin Stage when the meadow lay under a shallow wasting ice sheet. A stream of meltwater probably deposited the material in an embayment of the ice mass. When the ice melted, the deposit was left without supporting walls, and under the influence of gravity and rainwash it gradually assumed its present oval form. To the southeast of this deposit are two much smaller and flatter deposits of similar material. Both are still largely bare of vegeta— tion but are being encroached upon. The main “dune” bears no vegetation whatever but is surrounded by a carpet of xerophytes and a partial circle of small lodge- pole pine. East of the main dune, in a fringe of timber on a low rock ridge covered with Wisconsin moraines, is a small spur that is composed also of gravel and cobbles mixed with sand. This spur may be a record of the A35 glacial stream channel through which the sand, gravel, and coarser materials were washed out to the main “(111118.” GLACIATION OF THE UPPER KERN BASIN: KERN GLACIER SYSTEM As previously noted, the climax of glaciation in the Sierra Nevada occurred in the central part of the range. At the heads of the Merced and Tuolumne Basins, for instance, the cirques and their short outflow canyons were filled with Pleistocene ice nearly to their full depth; some were completely filled, and the ice over- flowed the dividing ridges and spurs on a large scale. Still farther northwest, in the rugged northern part of Yosemite National Park and the adjoining headwater areas of the Stanislaus and Mokelumne Rivers, the ice, even during the relatively moderate Wisconsin Stage of glaciation, covered the mountains except for a few of the highest peaks; The ice formed a local, flatly domed cap whose gently curving surface did not reflect the character of the topography underneath. Only isolated nunataks stood out above the ice. To any observer trained to read the signs of alpine glaciation, it cannot fail to be evident, as he travels southward from this area of maximum glaciation across the upper Merced, upper San Joaquin, and upper Kings River basins, that ice accumulation decreased by de- grees, that the cirques and canyons were filled to progressively less depth, and that the intermediate peaks and crests stood above the ice with greater height FIGURE 36.—Sand Meadows. These meadows and the neigh- boring Hockett Meadows are on a glaciated platform at an altitude of 8,500 to 9,000 feet in the Kaweah Basin. At the left, in front of the trees, is the so—called “sand dune,” in reality an outwash deposit. In the foreground, the meadow is strewn with glacial boulders. A36 and in more continuous chains. Nor did the ice in each of these basins possess a continuously domed surface; on the contrary, it had a broadly concave surface, each basin being an independent area of accumulation. In the upper Kern Basin, it will be clear from the foregoing, the paucity of ice, as compared with that in the other major basins to the north, was greatest of all. In spite of the great altitude of the upper Kern Basin— whose floor ranges for the most part from 10,000 to 11,500 feet—and in spite also of the great height of the surrounding peaks—Which range from 12,000 to more than 13,000 feet on the Great Western Divide and the Kings-Kern Divide, and from 13,000 to well over 14,- 000 feet on the main crest of the Sierra Nevada—the upper Kern Basin at no time during the Pleistocene Epoch, not even during the earlier stages of glaciation, contained an extensive unbroken ice sheet such as filled the upper Tuolumne Basin above the Yosemite region (Matthes, 1930, p. 77 and map, pl. 39). The glaciers that issued from the cirques on the surrounding crests converged toward the central canyon as distinct ice streams separated from one another by mountain spurs or low divides. Coalescence of these glaciers across interspaces occurred only in the uppermost part of the basin, but even there several spurs remained emergent above the ice. Nor did the ice attain great thickness. It averaged little more than 1,000 feet in depth on the broad plateaulike benches and reached maximum depths of 2,000 to 3,000 feet only in the main canyons. Systematic mapping of the individual branch glaciers showed that they filled the cirques to not more than two- thirds of their depth and filled the canyons leading out from the cirques to only about one-half of their depth. On the walls of those canyons, the upper limit of glacia- tion is in many places distinctly marked, owing to the fact that above that limit the walls are fluted directly downward by recurrent avalanche action, but below the limit they are smoothed by longitudinal glacial action. The paucity of ice thus revealed at and near the centers of ice accumulation surrounding the upper Kern Basin contrasts strikingly with the abundance of ice in comparable places farther north in the High Sierra. The scarcity was due not only to the far southern lati- tude of the basin and to its southward exposure but also to the effect of the Great Western Divide, which exacted a toll of precipitation from the moisture-bear- ing westerly winds and thus reduced the amount of snow available for distribution over the upper Kern Basin. Nearness to the deserts east and southeast of the Sierra Nevada undoubtedly was also a determining factor, during Pleistocene time, as at present, the hot SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY dry air from those areas causing much of the fallen snow to waste away by evaporation. That the ice in the upper Kern Basin constituted a separate system, which was independent of ice accumu- lations in adjoining drainage basins and was of strictly local origin, may be attributed to the following cir- cumstances: Only a few gaps in the encircling mountain crests were low enough to permit diversions of ice across them; furthermore, these few diversions were outgoing rather than incoming, owing to the great altitude of the upper Kern basin and the relatively deep dissection of the adjoining mountain areas—the Kaweah Basin on the west, the Kings Basin on the north, and the Sierra escarpment on the east. The term “Kern glacier system” may appropriately be used to apply, broadly, to all the Pleistocene ice bodies, large and small, of the upper Kern Basin. The specific members of that system are: First and foremost, the great Kern glacier itself—that is, the Kern trunk gla- cier and its many branches; the scattered little glaciers west of the canyon, at the sources of the Little Kern River and the Tule River; and, finally, those glaciers east of the Kern Canyon, also small, at the sources of Golden Trout Creek and the South Fork of the Kern River. In the following pages the various members of the Kern glacier system are discussed in the order in which they have just been named. By far the greater part of the system falls within Sequoia National Park but, in the interests of completeness, the parts of the system outside the park will also be considered briefly. Actually these outlying parts are, with but a few excep- tions, only a short distance from the park boundaries. KERN CANYON AND TRIBUTARY VALLEYS The Kern glacier was a great, many-branched ice body fed from ranks of cirques along the high border- ing ranges, the Great Western Divide (altitudes 10,000 to more than 12,500 ft), the Kings-Kern Divide (12,500 to more than 13,000 ft), and the main crest of the Sierra Nevada (13,000 to almost 14,500 ft), as well as on ridges subsidiary to these crests. Because the Kern Canyon extends in a nearly straight line through the middle of the upper Kern Basin and the tributary canyons branch from it like the ribs in an oak leaf from the main rib, the Kern glacier had much the same pattern. During the Wisconsin Stage, the tributary ice streams lay en- tirely confined in the side canyons, but during the El Portal Stage the ice spread locally across intervening divides and over the benchlands on either side of the main canyon. It has long been assumed that the Kern glacier never extended beyond the terminal moraine that loops across the floor of the Kern Canyon a mile south of Golden GLACIAL RECONNAISSANCE 0F SEQUOIA NATIONAL PARK Trout Creek. However, the present reconnaissance showed that this moraine marks only the limits which the glacier reached during the Wisconsin Stage and that during El Portal Stage the glacier extended about 7 miles farther down the canyon, to the vicinity of Hockett Peak. It appears, then, that during the Wis— consin Stage the Kern glacier attained a length of 25 miles, and during El Portal Stage a length of about 32 miles. In Glacier Point time, the glacier probably reached a length approximately equal to that reached during El Portal Stage. WISCONSIN STAGE KERN CANYON The lower limits of the great Kern trunk glacier of Wisconsin time are clearly indicated by the remnants of moraines in thevicinity of the Kern Canyon Ranger Station (pl. 1) . These moraines were first made known by Lawson (1904, p. 345—348). The author’s recon- naissance provided opportunity for examination of the moraines described by Lawson and for noting several others belonging to the same series but not previously reported. The author’s observations served to empha- size the close correspondence which exists between the Wisconsin moraines in Kern Canyon and the compar- able series in the Yosemite Valley (Matthes, 1930, p. 56—58). As was made clear by Lawson, the most southerly of the moraines is a V—shaped loop, whose apex is directed down the canyon, about three-quarters of a mile south- southwest of the Kern Canyon Ranger Station. The moraine is an inconspicuous one, being on the average only 25 feet high. Beginning at the west side of the canyon, this moraine runs out obliquely into a depres- sion bounded, on the east, by a linear rock ridge rising from the central part of the Kern Canyon, adjacent to the Kern River. This rock ridge is in the category which Lawson designated as “kernbuts.” After mak- ing a pointed loop in this depression, the moraine ascends to the top of the kernbut. Immediately north of the end moraine, and very similar to it, are two recessional moraines, also reported by Lawson, the northerly one of which is double crested. In crossing from the west canyon wall to the kernbut, these moraines also form V-shaped loops. As Lawson recognized, these moraines outline the positions of a small frontal lobe of the Kern trunk glacier which was thrust into the depression between the west canyon wall and the rock ridge. However, Lawson evidently did not observe that morainic mate- rial also lies on the steep east side of the rock ridge and that east of the Kern River there is at least one distinct moraine loop, standing 10 to 15 feet above an outwash terrace. This moraine must represent the eastern A37 equivalent of one of the three loops west of the rock ridge, probably the southerly one. Thus the rock ridge appears to have divided the glacier, at its terminus, into two small lobes. In this reconnaissance, several other indistinct mo- raines were noted a little farther upstream, in the area just south of Coyote Creek. Also, a prominent moraine , was traced from the Kern Canyon Ranger Station west. northwest along the north side of Coyote Creek as far as the lower west slope of the canyon. Where Coyote Creek debouches on the canyon floor, it is difficult to dis- tinguish the material of this moraine from that of the boulder levees along the stream. Still farther north, likewise on the west side of the Kern River (immedi- ately north-northeast of Soda Spring), are two other fairly prominent Wisconsin moraines, each the rem- nant of a moraine loop. Finally, outside the mouth of Golden Trout valley is the beautifully symmetrical moraine described by Law— son which is notable as the largest and most northerly member of the morainal series left by the Kern trunk glacier. This moraine also is V—shaped in ground plan and, having been breached at the apex by the Kern River, actually consists of two separate wings, one on either side of the river. These Wings are even-crested massive ridges, each almost half a mile long and in places rising nearly 100 feet above the flat along the river. This moraine may correspond to the one in Yosemite Valley which formed the dam of Yosemite Lake (Matthes, 1930, p. 57). Throughout the IVs—mile section of the Kern Canyon occupied by these moraines a jumble of material, both morainal and outwash, mantles the rock floor to an ap— preciable depth. What this depth may [be is suggested at a point a few hundred yards downstream from the ranger station where the river has cut through the mantle into the underlying granite (fig. 37). Here, on the east side of the river, the mantle does not appear to be over 30 feet deep, but up the valley, toward the upper moraine, the mantle may well gain in thickness. Northeast of Volcano Falls, the old Conterno Trail connecting the floor of Kern Canyon with the valley of Golden Trout Creek climbs the left lateral moraine of the Kern trunk glacier, reaching the top at 7,000 feet. This fairly conspicuous moraine at one time evidently formed a barrier across the mouth of the small hanging valley north of the ‘C‘onterno Trail, but the moraine is now crossed by a narrow gorge. The height of the moraine shows that the Wisconsin glacier at this point had a thickness of 600 feet plus the thickness of the glacial and alluvial material that now encumbers the valley floor—probably at least 50 feet. It follows that in a distance of nearly 2 miles the Kern trunk glacier A38 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY FIGURE 37,—Kern River, cutting in bedrock, near Kern Canyon Ranger Station at south border of Sequoia National Park. tapered down from a thickness of about 650 feet to its terminus. Farther up Kern Canyon, no moraines of the trunk glacier are found through a distance of more than 15 miles, but beyond that, in the head of the canyon, a fine series of lateral moraines is ‘present (Lawson, 1904, p. 354, 358). (These moraines and others mentioned on following pages are not shown on pl. 1.) Those on the east side of the canyon are “several miles in length and rise to an altitude of probably 1,500 feet” above the bot- tom of the canyon (Lawson, 1904, p. 354) ; they extend from just beyond the mouth of Wallace Creek north- ward beyond the head of the Kern Canyon into the canyon of Tyndall Creek. TRIBUTARY VALLEYS On the west side of the Kern Canyon, from Coyote Peaks to the Big Arroyo, the valleys tributary to the Kern are successively longer to the north, because the Great Western Divide, from which the tributary val- leys descend, diverges northwestward away from the northward-trending Kern Canyon. Because of this sit- uation and also because the Great Western Divide gains in altitude toward the northwest, the Wisconsin glaciers in these canyons were progressively larger to the north. The valley of Coyote Creek and also two little valleys heading just south of the Coyote Peaks held glaciers which were the most southerly ice bodies generated along the Great Western Divide; in fact, they were among the most southerly in the Sierra Nevada. These glaciers were not long enough to fill the entire length of their canyons and reach the Kern Canyon. However, ice streams in two of the branches of Coyote Creek did unite to form a very short trunk glacier. The promi- nent right lateral moraine of the north branch of this glacier can be seen from the trail to Coyote Pass. Far- ther up toward the pass, the trail crosses the right lateral moraine of a small separate glacier which descended from the cirque north of Coyote Pass. Considerably more impressive were the ice streams in the canyons of Laurel Creek and Rattlesnake Creek. These, like all the branch glaciers to the north of them, were tributaries of the Kern trunk glacier. T’hey headed in compound cirques on the Great Western Divide, and at the mouths of their canyons, which were hanging, they cascaded down to the trunk glacier in GLACIAL RECONNAISSANCE OF SEQUOIA NATIONAL PARK the Kern Canyon. The Laurel Creek glacier was 6 miles long, the Rattlesnake Creek glacier 8 miles long. The Big Arroyo glacier, almost 14 miles long, was the largest and most complex of the tributary ice streams in the entire Kern Basin; in fact, it constituted a major glacier system in itself. Most of the sources of this glacier lay along the Great Western Divide, in a 15- mile rank of capacious cirques, but some lay to the north— east, in cirques of the Kaweah Peaks Ridge. As a result of severe glacial erosion, Big Arroyo Can- yon, like Lyell Canyon in Yosemite National Park, has a smoothly U-shaped cross section, and the canyon shoulders are rounded from overriding by the main glacier (figs. 38, 39). Upper reaches of the canyon, which are mostly devoid of timber as a result of snow- slides, are broadly open. Many of the cirques along the Great Western Divide are remarkably smooth- floored and clear of debris; the little debris that is pres- ent is usually concentrated in a few narrow belts. In these cirques lie numerous scattered lakes and chains of lakes. These erosional features are strikingly visible from the High Sierra Trail, which crosses Kaweah Gap and descends along the left side of the Big Arroyo to Kern A39 Canyon. So also are some of the Wisconsin lateral moraines of the main glacier and its tributaries. The sloping platforms at the base of the Black and Red Kaweah Peaks, for example, are littered with moraines and erratics, mostly above timberline. The most interesting display of moraines is to be found on the southwestern part of the Chagoopa Pla- teau, which is also traversed by the High Sierra Trail. Moraine Lake, in this area, is one of the few wholly moraine-impounded lakes in the Kern Basin (figs. 4-0, 41). The little valley which the lake occupies was in- vaded by a side lobe of the Big Arroyo glacier. This lobe left. a series of concentric moraine loops, the last- formed and smallest loop being one at the lower end of Moraine Lake (the impounded water seeps out from the lake through this embankment). West of Sky Parlor Meadow the concentric moraines are separated by strips of meadow formed where sand has been washed down from the flanking moraines. The outer- most moraine forms the west boundary of Sky Parlor Meadow. This moraine, when followed to the south- east, reaches the crest of a ridge and here makes an abrupt turn toward the southeast; it then continues parallel to the general course of the Big Arroyo glacier. FIGURE 38.—-View from the vicinity of the Nine Lake Basin southward down Big Arroyo, which, in the distance (left), becomes a deep U-shaped canyon flanked by forested plateaus. Photograph by W. L. Huber. A40 FIGURE 39.—View across Big Arroyo from an unnamed moun- tain east of Little Clair Lake. Big Arroyo, like the Kern Canyon, is a stream-cut canyon modified by glacial action. The broad peak to the right of center is Mount Kaweah. On the far side of Big Arroyo is part of Chagoopa Plateau. The moraine, when followed north along the edge of Sky Parlor Meadows, describes a big arc. In this sec- tion it has several subsidiary crests, and its steep front increases in height to 100 feet or more. Sky Parlor Meadow follows around the arc of the Wisconsin mo- raine and gradually narrows into a point. Beyond it lie the smooth older moraines. The westerly tributary glaciers which remain to be SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY considered ranged from 3 to 8 miles in length and there- fore were small as compared with the Big Arroyo gla- cier. These ice streams headed along the Kaweah Peaks Ridge or its subsidiary spurs, with the exception of the Kern-Kaweah glacier, which originated along three dif- ferent crests: the Kaweah Peaks Ridge, the Great Wes- tern Divide, and the Kern Ridge. The several small glaciers on the southerly slopes of Mount Kaweah and the Red Spur descended onto the northern part of the Chagoopa Plateau; here they coalesced to form a small compound ice sheet which discharged southeastward into Kern Canyon through the hanging valleys of Red Spur Creek and the next stream to the south. The Kern-Kaweah glacier had broad, capacious c.01- lecting basins which were among the largest of the en— tire Kern glacier system, and it was therefore a major aflluent of the Kern trunk glacier. It will be recalled that the Kern-Kaweah glacier was one of the three ice streams whose union formed the head of the Kern trunk glacier. The glacier formed the apex of the Kern Canyon glacier system, in that it lay in line with the trunk glacier at the very head of the main canyon; it was a FIGURE 40.—Chagoopa Plateau and Moraine Lake, viewed from the edge of Big Arroyo. The plateau and timbered benches on the far side of Big Arroyo are remnants of an erosion surface left after trenching of the canyon. GLACIAL RECONNAISSANCE 0F SEQUOIA NATIONAL PARK FIGURE 41.—Moraine Lake, on the Chagoopa Plateau, one of the few wholly moraine-enclosed lakes in the Kern Basin. water seeps out through a morainal embankment at the lower end of the lake. complex ice body about 8 miles long and almost as broad. This glacier was the central member of the three whose union gave origin to the Kern trunk glacier. The many cirques of this central glacier lay on the four crests, Kern Ridge, the Great ‘Vestern Divide, the Kings—Kern Divide, and the main Sierra crest, which formed a lofty rim encircling the Kern glacier on all sides except the southwest. ‘Vhere the canyon “steps up” from main Kern Canyon just above the junction of the Kern-Kaweah River, a moraine loop of this glacier is crossed by the trail. The river has cut through the moraine and into the underlying bedrock; it thus gives rise to a series of falls. More conspicuous evidence of the Wisconsin glaciation occurs at higher altitudes, where there are 60-odd lakes along the upper stream courses and Within the cirques of this basin as well as a multitude of smoothly rounded ledges among which the trail precari- ously makes its way up to the sources of the Kern River. Many of the knolls and ridges, curiously, reverse the relationship typical of ordinary roches moutonnées in that they are more abrupt on the upvalley side than on % «whwlwaM‘E‘ The Photograph by W. L Huber. the opposite side. Their anomalous profiles reflect the control of two joint sets, one dipping rather steeply up- valley, the other dipping less steeply downvalley. The Tyndall glacier deposited a series of bouldery left lateral moraines which the John Muir Trail crosses in its course south of Tyndall Creek. One of these moraines holds in the long, narrow lake west of the long spur extending southwest from Mount Tyndall. The moraines are not very prominent, except for the top- most one which is a barren, sharply defined ridge con- trasting rather sharply with the smooth bare slopes of the Bighorn Plateau above it (fig. 42). The three easterly tributaries of the Kern trunk glacier—VVallace glacier, Whitney glacier, and Rock glacier—all came from the highest of the crests border- ing the Kern basin—that is, from the main Sierra divide, which in this section bears several 14,000-f00t peaks and many summits which are but little lower. lVallace glacier, about 7 miles in length, was joined by a major branch from the northeast, the Wright glacier. The numerous moraines of this glacier sys- tem are bouldery and regular (fig. 43), and some are A42 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY FIGURE 42,—View up the valley of Tyndall Creek from the Bighorn Plateau, which appears in the lower right fore- ground. The flat-topped peak to the right of center is Diamond Mesa. The barren ridges of rock debris extend- ing toward the lower left-hand corner of the photograph are lateral moraines of the Wisconsin Stage. The upper- unusually prominent. The John Muir Trail traverses them in its course across the basin, above the junction of Wright and Wallace Creeks. There are at least four right laterals and two left laterals of the Wright glacier, and three left laterals and two right laterals of the upper Wallace glacier. A medial moraine given off from the end of the ridge between Wright and Whitney Creeks extends west-southwestward down the valley for about a mile, as far as Wright Creek. Wallace, Tulainyo, and Wales Lakes, in upper Wal- lace Creek basin, are glacial lakes of exceptionally large size, each being more than half a mile long. Wallace Lake, which is probably at least 100 feet deep, lies in a great compound cirque; it marks the confluence of two ice streams, one from a cirque on Mount Barnard, the other from a larger, higher cirque under the summit of Mount Russell. Tulainyo Lake (fig. 44) lies in the latter cirque. The environment of both lakes, being above timberline, has a grim, barren aspect. Tulainyo Lake (altitude 12,856 ft), said to be the highest body of water in North America, occupies a cirque that doubtless held a small glacier until very recently. The lake is enclosed by a massive moraine through most moraine marks the highest level reached by the Tyn- dall glacier during that stage. The lower, timber—covered moraines were laid down during the recession of the glacier. The smooth slope of the Bighorn Plateau seems devoid of glacial features, yet in places it bears scattered erratic boulders left from El Portal Stage. (See fig. 26.) which the water seeps out. Wales Lake lies in a short tributary canyon which, from the evidence of glacial sculpturing on its walls, held an ice tongue about 1,000 feet deep. On the granite around Wallace Lake, surprisingly little glacial polish remains, and in some places weather pits have been formed in the surfaces of hori- zontal platforms. This condition implies very rapid postglacial disintegration as compared with that ob- served in the Yosemite region. Half a mile below Wallace Lake is a striking exhibit of glacially quarried rock; several sheer rock faces are determined by vertical master joints. The clifl’ be— low Wales Lake also exhibits glacial plucking and abrasion guided by joints. Where the John Muir Trail crosses Wallace Creek, the stream through a stretch of 400 feet has trenched the bedrock to a depth of 25 to 50 feet (fig. 45). The gorge has vertical walls along which stand tottering joint slabs. The stream cutting here, which must be wholly postglacial, contrasts vividly with the slight erosion effected by the Merced River above Vernal and Nevada Falls (Matthes, 1930, p. 69). The differ- GLACIAL RECONNAISSAN CE 37? 0F SEQUOIA NATIONAL PARK A43 FIGURE 43.—View northeastward up the valley of Wright Creek toward Mount Tyndall (central peak with gullied slopes). Stretching across the valley floor are two moraines that mark brief halts in the recession of the Wright glacier. In ence may be attributed to the fact that, at the gorge, Wallace Creek is cutting into granite having numerous intersecting joints which divide the rock into blocks and slabs of moderate size, whereas above Vernal and Nevada Falls the Merced is flowing over massive gran- ite. Wherever the granite is sparsely jointed or wholly undivided over considerable distances—and there are many such places in the High Sierra—postglacial stream erosion can be measured in inches rather than feet. Along Wallace Creek, less than 300 yards to the north of the gorge the same granite, where sparsely crossed by joints (10—20 in. apart), forms a smooth, untrenched valley floor. The Whitney glacier was a forked ice body, its south branch, the Crabtree glacier, being almost as long as the upper part of Whitney glacier itself. Thus this glacier system somewhat resembled the Wallace system but was smaller and only 6 miles long. It is noteworthy that the sources of this glacier were on that lofty section of the Sierra crest that includes Mount Whitney (14,495 ft), the highest peak in the con- tinental United States outside of Alaska (figs. 10, 46). From the glacial sculpturing and avalanche fluting on its valley walls, described below, the Whitney glacier seems to have been about 1,200 feet thick in its upper reaches and 900 to 500 feet thick in its lower valley. On both sides of the valley, below the fork, the trail the canyon to the right of Mount Tyndall, the level at which the surface of the glacier lay is clearly defined by the lower limit of the gullies cut in the cliff. These gullies are due to avalanche erosion. from Sandy Meadow to Guyot Flat crosses several lateral moraines of the Whitney glacier. The left laterals, composed in large part of big angular boul- ders, are exceedingly rough. In places the trail fol- lows sandy strips between the moraines. - Upper Whitney Creek, like upper Wallace Creek, has entrenched itself in the canyon floor to depths of about 20 feet in those places where stream erosion has been facilitated and directed by vertical master joints. In Upper Whitney and Crabtree Canyons, the floors and sides are irregular as a result of selective glacial quarrying (figs. 47, 48) ; the sheer headwalls strikingly exhibit avalanche sculpture (Matthes, 1938, 1950a). The north wall of Mount Hitchcock, for example, is deeply cleft by numerous sharply incised recesses, a special form of avalanche chutes cut across a system of well-formed vertical joints (fig. 49). Debris cones, which were still covered with snow when observed, give evidence of continued avalanche action. The east face of Mount Hitchcock, on the other hand, has deeply cut avalanche chutes that have been controlled ‘by ver- tically sheeted structure in the rock (fig. 50). The southwest flank of Mount Young is rather intricately sculptured by avalanches, and, between the chutes, the ribs stand out as sharp pinnacles. Parts of the west side of Mount Muir and the crest extending south from SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY A44 an ananwouona 8E4 .woafimmc an... E 5% 3 hawk? £530 4535 MdaHrm 203 93 Bob 33> US$52 «.3wa 9: we $0.5 BE: 05 no :ME .333 omflwal. 3“ 953% GLACIAL RECONNAISSANCE OF SEQUOIA NATIONAL PARK A45 FIGUBE 45.-—-Gorge 50 feet deep, cut by Wallace Creek, for the most part in postglacial time. this peak have been subjected to avalanche erosion; that part of the ridge to the west of Whitney Pass has sharp ribs between chutes. The trail laid across these feas- ures is in a precarious position and is inevitably subject to being swept away periodically by avalanches. Some of the ribs are particularly frail and picturesque. The long spur south of Crabtree Creek is also avalanche scarred on the north side. A small independent glacier that descended the north slope of Mount Guyot left records which attest to its great eroding power. The cirque of this glacier is shal- low but unmistakable, and a bouldery outer moraine in- dicates that this little glacier turned west and, at an altitude of 10,000 feet, a mile below its source, ended in a narrow tongue. The Rock glacier was the most southerly and the longest (11 miles) of the tributaries which joined the Kern trunk glacier from the east. The moraines of the Rock glacier cloak both slopes of the lower valley, and several small moraines forming partial loops occur on the bottom of the canyon. Intermediate spaces consist, for the most part, of wet, boggy meadow. In the lower valley of Guyot Creek, moraines outline a small Wis- consin glacier which fell short of joining the Rock glacier, and at the heads of two tributaries of the South Note man standing on the right brink. Stream erosion here has been facilitated by the numerous intersecting joints. Fork Rock Creek, southwest of Siberian Outpost, there are records of two other, even smaller, Wisconsin glaciers. EL PORTAL STAGE On the basis of the characteristics of the glacial de- posits of El Portal age as set forth on previous pages, the deposits of this early stage in the Kern Canyon and in its branches may now be described. To the author it was soon apparent that El‘Portal moraines in the Kern Canyon at the junction of Coyote Creek are situated with respect to the Wisconsin mo- raines precisely as are El Portal moraines in the Yosem- ite Valley. In that valley the author discovered, in 1913, that at a point 1 mile below the terminal moraine of the Wisconsin Stage the highest left lateral of El Por- tal Stage lay 2,200 to 2,300 feet above the valley floor (Matthes, 1930, p. 66—68). A few miles farther down the valley, the right lateral was at a corresponding height. Once identified, both laterals were readily traced down the Merced Canyon to the vicinity of El Portal. In the Kern Canyon, similarly, the lateral moraines of El Portal Stage lie at heights ranging from 1,600 to 2,000 feet above the terminal moraine of the Wisconsin Stage. A46 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY FIGURE 46,—Mount Whitney, viewed from the west. The precipitous clifis of the mountain, the scoured bedrock floor of the canyon, and the small lake in the foreground are typical features of the glaciated upper Kern Basin. The cliffs are furrowed by avalanche chutes. Search for these older glacial deposits was begun on the west side of the Kern Canyon. The upper zigzags of the trail that leads to the hanging valley of Coyote Creek were cut into a large body of El Portal mineral; ‘they thus afl'ord abundant exposures of the limonite- coated boulders in the material. The mass reaches all the way up to the mouth of the hanging valley, but inasmuch as this mass connects there with old deposits of the tributary Coyote glacier, it was deemed advisable to extend the search northward along the shoulder of the canyon. There, patches of the mass were found on a part of the upland where none could have been con- tributed by either the ancient Coyote glacier or the ancient Laurel glacier, the next tributary ice stream to the north. This old glacial material is for the most part hidden from view by a mantle of forest soil and granite sand, but cobbles of characteristic El Portal aspect occur in the holes left by uprooted trees. The upper margin of the deposit is ill defined but seems to range from 8,000 to 8,100 feet in altitude. It therefore lies 1,600 to 1,700 feet above the canyon floor; and, because the canyon floor is covered with late-glacial and alluvial deposits to a depth of probably 100 feet, the total thick- Photograph by W. L. Huber. ness of the ancient Kern glacier indicated is about 1,700 to 1,800 feet. The search on the east side of the canyon gave even better results. The old 'Conterno Trail to the upland valley of Golden Trout Creek first. ascends the left lat- eral moraine of the Wisconsin Stage. The trail sur- mounts the sharp bouldery crest at an altitude of about 6,900 feet and then winds up the gully between the crest and the mountainside as far as the mouth of a small hanging valley at an altitude of 7,000 feet. The lateral moraine at one time doubtless formed a barrier across this hanging valley, but it has since been notched by the streamlet. In the hanging valley itself, which is little more than a recess in the canyon wall, the topographic forms are not in the least suggestive of glacial action. They consist of alternating spurs and ravines converg— ing toward the central drainage line; yet these features, clearly because of normal subaerial weathering and stream erosion, are carved largely from glacial deposits of El Portal age. The entire recess is lined with such material, and because there is no evidence of this recess having contained a small tributary glacier, the deposits in question can be attributed only to the Kern glacier GLACIAL RECONNAISSANCE OF SEQUOIA NATIONAL PARK A47 FIGURE 47.—Junction of Crabtree Canyon (foreground) and Whitney Canyon (background). The sparsely jointed floor of Crabtree Canyon shows the effects of glacial quarrying. The intercanyon ridge in the middle distance has angular itself. In fact, several obscure lateral moraines of that glacier, arranged in tiers one above another, like ter- races, are discernable. The highest, which lies at an altitude of about 8,500 feet, shows that the Kern glacier of the El Portal Stage here filled the main canyon to a depth of fully 2,000 feet above the present floor; and, inasmuch as this floor is underlain by an estimated 100 feet or more of late-glacial and alluvial deposits, the total depth of ice may have amounted to 2,100 feet. The Conterno Trail, after making several zigzags, turns southward toward the valley of Golden Trout Creek. Along this stretch of the trail only occasional cobbles of El Portal age are found, the canyon sides being too steep to retain much loose material. But in the saddle behind the projecting crag marked 8,119 feet on the topographic map, a small body of El Portal material remains preserved (fig. 51). This body lies fully 500 feet above the adjacent part of the valley of Golden Trout Creek, and, inasmuch as the ancient Kern glacier doubtless rose at least 200 to 300 feet higher, hackly forms produced by glacial quarrying in well—jointed granite. The Whitney glacier spilled over the ridge and quar- ried its downstream side. The well-jointed canyon slopes in the background also exhibit the effects of glacial quarrying. there seemed good reason to believe that a side lobe of that glacier had pushed up into the valley. A rapid examination of the valley, indeed, showed that such had happened. Although the farthest limits reached by the lobe are not marked by a distinct moraine, granitic boulders left by the ice lie scattered on the surface of the basalt flow that fills the bottom of the valley. Along the southern edge of the basalt flow, at an alti- tude of about 8,100 feet, there is also a continuous moraine composed of a mixture of basaltic and granitic boulders. That moraine extends about 1 mile up the valley and indicates the approximate length of the ice lobe. At the mouth of the valley the moraine is indis- tinct, but it can be traced to an altitude of nearly 8,300 feet. Here, then, the ancient Kern glacier attained a thickness of about 1,900 feet in the main canyon. This measurement accords well with that obtained in the recess to the north of Golden Trout Creek, for the ancient glacier must have decreased in thickness down- valley in this part of its course. A48 SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY FIGURE 48.—View from the Whitney Trail across upper Whit- ney Canyon and Hitchcock Lake (foreground) at Mount Hitchcock (center). Zones of fracturing in the floor of the canyon have controlled the erosive action of the glacier. Lakes and snowdrifts mark the depressions quarried out in these zones. The intermediate humps, composed of sparsely Thus the great depth of ice in the Kern Canyon that is consistently indicated at three different points im- plies, of course, that the glacier of El Portal Stage extended several miles beyond the terminal moraine which Lawson believed marked the southernmost limit of glaciation. A methodical search for remnants of lateral moraine of El Portal age in recesses in the can- yon walls was clearly called for, but because the time available did not permit such a search, it was deemed best to examine the area about the southern end of Grasshopper Meadow, where the canyon contracts from its broad U-shape to a narrow V-shape and where, pre- sumably, glacial erosion had been largely supplanted by deposition under the thinning sluggish terminal part of the glacier. There, indeed, a considerable body of El Portal moraine was discovered to the east of the trail, in the reentrant through which the trail ascends to the col at the head of the defile called Trout Meadows. fractured rock, have been subject chiefly to the slower process of grinding. Except for the accumulation of rock debris at the base of Mount Hitchcock, this part of the canyon has undergone only insignificant changes since the disappearance of the glacier. Flewelling. Photograph by Kenneth The deposit is partly covered by soil and forest litter, but the cobbles in it are plainly visible on the slope to the east of a sharp-cut gully. The upper limit was determined, by aneroid, at an altitude of about 6,600 feet—that is, at a height of 900 feet above the bed of the Kern River, according to the contour lines on the topo— graphic map. If it be assumed that the river has deep- ened its bed 200 feet since El Portal time—observations made in other canyons farther north in the Sierra Nevada indicate that a greater amount is not probable—— the ancient Kern glacier at this lower point still was about 700 feet thick. The glacier may be reasonably supposed, therefore, to have penetrated at. least 2 miles farther down the canyon, and the ice front, accordingly, must have lain at an altitude of approximately 5,700 feet about 7 miles south of the terminal moraine of the Wisconsin Stage—that is, at a place locally called “Hole-in-the-ground” (name not indicated on topo- GLACIAL RECONNAISSANCE 0F SEQUOIA NATIONAL PARK A49 FIGURE 49.—N0rth side of Mount Hitchcock, viewed across Whitney Canyon. An unusually fine series of parallel avalanche chutes is here shown. These chutes have been formed across a system of vertical joint fractures in the granite; their positions are not determined by master joints graphic map), which is within the bend of the canyon north of Hockett Peak. Inasmuch as the morainal deposit has been traced to a point only 50 feet below the col at the head of the Trout Meadows defile and may not truly indicate the highest level attained by the ice, it seemed possible that for a brief period the ice had risen high enough to spill through the col. Search for morainal material was therefore made in the upper part of Trout Meadows, but none was found. It was concluded, therefore, that the Kern glacier of El Portal Stage had remained con- fined to the canyon. On the way down the canyon, and on the way back, a lookout was kept for El Portal deposits. Only scattered boulders and occasional small patches of cobbles were observed on the canyon floor. The large kernbut to the south of Kern Lake was also examined for old glacial material because it was realized that if this kernbut, which stands only about 300 feet high, was in existence at the time of El Portal glaciation, it would have been overtopped by the ice. No glacial material was found on the main summit, which is directly east of the col The chutes all terminate at the upper limit of glacial action, below which the can- yon wall is straight, although it is hackled in detail by glacial sculpturing. extending parallel to their axes. that is traversed by the trail, but on the somewhat lower crest to the west of the trail there is a roundish boulder of granite nearly 3 feet in longest diameter. That boul- der very probably is a glacial erratic; it owes its round- ish shape not to glacial abrasion but to the spalling off of thin curving shells, some of which lie at its sides. A few feet away is a fragment derived from it that has an exposed part measuring 23 inches in length, 18 inches in breadth, and 9 inches in height. Boulders of pre-Wis— consin age in the Sierra Nevada commonly break up in this very manner. The boulder in question is composed of a fine-grained even-textured granite, as yet unstained by ferric oxides; however, the local rock, which crops out on the crest, only 10 feet away, consists of a coarse— grained and obscurely porphyritic granite stained a ruddy tint as the result of the oxidation of its ferro- magnesian minerals. Nor was any fine—grained granite, such as that of this boulder, discovered elsewhere in the vicinity. That the boulder is foreign to the locality and was dropped on the crest by El Portal ice, therefore, seems a reasonable presumption. A50 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY FIGURE 50.—East face of Mount Hitchcock. The entire moun- tain has a vertically sheeted structure, and infiltration of water and consequent frost action are facilitated along the weaker zones. There the rock is split into thin plates and To the south of this kernbut, the trail hugs the west side of the canyon (the location of the trail is not shown correctly on the topographic map), winding its way among the crags and blocks that have been brought down by rockslides. All this slide debris is angular and is gray in tone, but among the debris lie rounded rust- hued boulders and cobbles that are typical El Portal material. They must be derived from an old lateral moraine high up on the canyon side. Such boulders and cobbles are more abundant farther on, in the fan built by the streamlet that empties into Little Kern Lake. These doubtless have been washed down from the older moraines of the small hanging glacier that headed under the Coyote Peaks. Particularly significant is the presence of El Portal material on the north slope of the next great kernbut—— the one immediately south of Little Kern Lake. There is nothing to suggest that this material has been dis- lodged and redeposited; it appears to lie in the place where the Kern glacier left it. A thorough search for morainal material on the top of the kernbut could not be made in the time that was available. None was found, but it is reasonably certain, nevertheless, that El Portal ice passed over the kernbut, slivers; the fragments loosened by frost are then swept down by avalanches. These chutes stand in marked contrast to those shown on the frontispiece, which are not controlled by fractures but are worn in massive granite. because interpolation between the altitude, 8,300 feet, of the right lateral moraine south of Coyote Creek and its altitude, 6,600 feet, in the recess near the Trout Meadows col indicates that at the kernbut in question the surface of the ice lay at about 7,300 feet—that is, fully 300 feet higher than the top of the kernbut, which, according to the topographic map, has an altitude of 6,981 feet. South of the kernbut, in the wide space of Grasshop- per Meadow, the canyon floor is aggraded with large quantities of coarse angular rock debris derived from the toe of the great rockslide on the west side of the canyon. Some of this debris has recently been cut away by the river, and as a result there is now a flight of five clean-cut terraces about 50 feet in aggregate height; but the glaciated floor of the canyon and whatever morainal material may rest on it remain hidden from View. No attempt was made to determine the farthest limits which El Portal ice might have reached in the Kern Canyon, previous search in other canyons on the west slope of the Sierra Nevada having demonstrated that as a general rule the trunk glaciers of El Portal Stage have left no recognizable terminal moraines and GLACIAL RECONNAISSANCE OF SEQUOIA NATIONAL PARK A51 FIGURE 51.——View eastward across the Kern Canyon in the vicinity of Kern Canyon Ranger Station. The light-colored spur to the left is granite; the gray tonguelike mass to the right is a lava flow of basalt that cascaded down from the valley of Golden Trout Creek. At the near edge of the lava that, for this reason and because of the erosional changes that have taken place in the canyons, the far- thest limits attained by those ancient glaciers are not definitely marked. Besides, it seemed probable that the aggraded conditions noted in Grasshopper Meadow would extend some distance farther down the canyon and that but little loose glacial debris would find lodge- ment on the steep canyon walls. It remains to note the occurrences of El Portal mo‘ raine at scattered localities within some of the valleys tributary to the Kern Canyon and on the interfluves between these valleys. One of these deposits, at the mouth of the hanging valley of Coyote Creek, has al- ready been mentioned. Farther north, the Chagoopa Plateau is littered with the relatively smooth moraines of this age, which extend across Sky Parlor Meadow and tothe southeast of it. Near the head of the Kern River basin, the smooth, bare slopes of the Bighorn Plateau, between Tyndall Creek and Wright Creek, though seemingly devoid of flow, Golden Trout Creek has incised a deep, slotlike gorge in the basalt and even in the granite underneath. At the upper left is a rocky promontory; in the saddle behind it, 1,500 to 1,600 feet above the floor of Kern Canyon, occurs moraine of El Portal Stage. glacial features, are covered with formless older mo- rainal material, the disintegration of which has pro- duced the abundant sand on the plateau. In places the plateau bears scattered erratic boulders, many of which are in process of being broken up. One large erratic, south of the nameless pond in the middle of the plateau, measures 16 by 6 by 7 feet (fig. 26). It is composed of bluish granodiorite containing abundant biotite and hornblende crystals and is obscurely porphyritic. In these respects it contrasts with the local granite, which is conspicuously porphyritic, with phenocrysts or ortho- clase 2 to 3 inches in length. El Portal moraine also veneers several other wide tracts lying east of the Kern Canyon: the Sand Meadows region, between Wallace Creek and Whitney Creek, the broad upland south of the mouth of Whit- ney Canyon, and both sides of the lower valley of Guyot Creek. The deposits northwest of Guyot Creek are doubtless remnants of the right lateral moraine of the earlier Guyot glacier; those deposits southeast of the A52 creek probably are composite, including both the left lateral of the Guyot glacier and the right lateral of the Rock glacier. The deposits are very sandy, except where the slope is sufficiently steep to permit rainwater to wash away the sand; there many boulders are exposed. GLACIER POINT STAGE In only three localities in the upper Kern Basin were any glacial deposits observed that might with some assurance be assigned to the Glacier Point Stage (see pl. 1). Two of these localities are on the Chagoopa Plateau and permit no safe inferences regarding the depth which the Glacier Point ice may have attained.4 The third locality is 2 miles north of Golden Trout Creek, and the deposit there, fortunately, is of such character and so situated that there can be little doubt that it belongs to the Glacier Point Stage and marks ap— proximately the highest level attained by the ice of that stage. The deposit last mentioned lies on the otherwise bare level summit platform of the rocky knob that stands north of the little hanging valley previously mentioned (page A46). The material is very scanty, consisting of some cobbles and pebbles of granitic rocks. These rocks, however, have, with one exception, subangular forms such as are produced characteristically by glacial action. The exception is a smoothly rounded elongate pebble that is unquestionably stream worn; but such pebbles are bound to occur here and there in glacial deposits. All the cobble-s and pebbles are deeply stained by ferric oxides and break readily under the hammer. In general appearance these deposits do not differ from morainal material of El Portal age, but their position at an altitude of 9,000 feet, 500 feet above the highest El Portal moraines in the little hanging valley nearby, would seem to preclude the possibility of their belonging to that series of moraines. That these cobbles and peb- bles do not look more deeply decayed than El Portal material is probably due to the fact that they lie on an almost clean platform that is well drained and thor- oughly insolated, and where, consequently, the con- ditions are unfavorable for the action of chemical proc- esses of rock decay. The position of this ancient glacial material on the rocky knob implies that the ice of the Glacier Point Stage spread laterally over the upland to the east of 4These two localities are both indicated on Matthes’ maps, but only the southerly one, southeast of Sky Parlor Meadow, is mentioned in his field notes. According to these notes, this southerly locality is on a small knob of diorite which stands perhaps 60 to 70 feet above the level of the surrounding surface. The knob bears several erratics of siliceous granite of dimerent types. One erratic 4 feet in diameter, near the summit of the knob, is exfoliating. F. F. SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY the canyons. The ice may have invaded the upland for a distance of three-quarters of a mile, as far as the 9,000-foot contour line shown on the topographic map. There should consequently remain some patches of Glacier Point material here and there on the upland and, likewise, vestiges of a lateral moraine in the vi— cinity of the 9,000—foot contour line or even somewhat higher up. For such vestigial remnants, however, the author was unable to make a yard—by—yard search in the time that was available for the reconnaissance. It is hoped that others who may be interested in the problems of the Kern Canyon will someday make a thorough search on the upland surrounding the rocky knob and as far south as the valley of Golden Trout Creek. Inasmuch as the rocky knob has an altitude of about 9,000 feet, the glacial deposit on it lies fully 2,500 feet above the present aggraded floor of the Kern Canyon and, therefore, probably as much as 2,600 feet above its glacially excavated rock floor. These relations, how— ever, do not imply that the Kern glacier of the Glacier Point Stage here attained a maximum depth of 2,600 feet, for during that early stage the canyon had not yet been excavated to anywhere near its present depth (as measured to the rock floor)——most likely 400 to 500 feet greater now at the locality under consideration than during the Glacier Point Stage. The glacier therefore was presumably only about 2,000 feet thick; but if so, it is noteworthy that this glacier was as thick as its successor of El Portal Stage. It does not follow, how- ever, that the glacier was either as long or as powerful, for during Glacier Point time the canyon doubtless had not yet been much enlarged from its narrow stream- worn V-shape and therefore offered a much less perfect channel for the ice to flow through than later, when the canyon had acquired a capacious U-shape. Nor did the Kern glacier of the Glacier Point Stage possess as great kinetic energy as its successor of the El Portal Stage, for a larger proportion of its mass was diverted laterally over the uplands and therefore did not take part in the organized flow movement of the ice stream within the canyon. Yet, in spite of these adverse cir- cumstances, the Kern glacier of the Glacier Point Stage probably extended down the canyon several miles be- yond the terminal moraine of the Wisconsin Stage. The farthest point attained by this glacier remains, for the present, a matter of conjecture, but, to judge by the rates at which other trunk glaciers of comparable magnitude on the west slope of the Sierra Nevada appear to have decreased in thickness toward their termini, it seems not unlikely that the Kern glacier of the Glacier Point Stage reached at least as far as the upper end of Grass- hopper Flat, perhaps even as far as the lower end. GLACIAL RECONNAISSANCE OF SEQUOIA NATIONAL PARK SUMMARY STATEMENT The upper Kern Basin was occupied, in the Pleis- tocene Epoch, by the most southerly of the great glacier systems of the Sierra Nevada. Being less favorably situated, in certain respects, than its principal neigh- bors to the north, the Kern glacier was of smaller volume, but nevertheless it attained impressive size during each of the three recorded glaciations: the Glacier Point Stage, El Portal Stage, and the Wisconsin Stage. Banks of cirques on the Great Western Divide, the Kings-Kern Divide, and the main crest of the Sierra Nevada fed the many branch glaciers which, in impos— ing succession, joined the trunk glacier in the main canyon. The maximum extent reached by the Kern glacier during the earliest, or Glacier Point Stage, cannot be determined with certainty; however, the fact that in the region above Golden Trout Creek the glacier at- tained about the same thickness as its successor of the second, or El Portal Stage, would seem to warrant the inference that it advanced approximately the same dis— tance—that is, to a point in the Kern Canyon about 7 miles south of the park boundary. The record of El Portal Stage, though incomplete and decipherable only with difficulty in many places where it is preserved, can be spelled out with much greater assurance. The distribution of El Portal gla- cial deposits in and about the Kern Canyon indicates that the ancient Kern glacier of this stage, though in most respects similar to its successor of the latest, or Wisconsin Stage, attained greater volume of ice and, accordingly, greater thickness and length. There being more ice than the canyons could hold, in some places, the ice locally spread across intervening divides and over the benchlands on either side of the main canyon to a total breadth of 4 to 6 miles; thus a central ice sheet of about 30 square miles was produced—a remark— able fact considering that the entire expanse sloped southward and lay exposed to the rays of the midday sun. The trunk glacier extended fully 7 miles beyond the limit reached in the Wisconsin Stage»#that is, it advanced 7 miles south of the boundary of Sequoia National Park—the evidence indicating that the termi- nus of the glacier, when farthest down the canyon, lay at an altitude of 5,700 feet in the bend to the north of Hockett Peak, at lat 36°14’ N. (This latitude may represent the most southerly point reached by glacial ice in the Sierra Nevada.) The overall length of the Kern glacier system was then 32 miles. The thickness of the trunk glacier is reliably indicated at several points where its gradually thinning lower portion once rested; this thickness ranged from about 2,100 feet at A53 the mouth of the small hanging valley just north of Golden Trout Creek to 700 feet at the entrance to the Trout Meadows defile. In the Wisconsin Stage the Kern glacier system re- mained a sprawling ice body whose trunk and branches lay confined within their respective canyons as distinct ice streams separated from one another by mountain spurs or low divides. The overall length of the system was 25 miles. Seven of its tributaries were 6 to 11 miles long, and the Big Arroyo tributary was 15 miles long. At the mouth of Wallace Creek canyon, the trunk glacier was fully 2,700 feet thick; and where joined by Whitney glacier, 2 miles farther downstream, its thickness was still almost undiminished, being be- tween 2,600 and 2,700 feet. Thence, however, its thick- ness declined rapidly toward the terminus, being about 1,900 feet at the Big Arroyo glacier, 1,600 feet at the Rattlesnake glacier, 650 feet at the mouth of the hang- ing valley just north of Golden Trout Creek, and 200 feet opposite Tower Rock. The farthest point reached by the terminus of the trunk glacier coincides with the south boundary of Sequoia National Park, the boundary posts of the park standing, at an altitude of 6,350 feet, on the curving outer moraine marking the extreme limits of the Wisconsin glaciation. The trunk glacier, so much more powerful than its confluent tributaries, excavated its trough hundreds of feet below the level of their canyons, leaving them hanging. The many side valleys joining the Kern Canyon farther south have also been left hanging. Doubtless they became so in the first place in preglacial time as a result of the rapid trenching of the Kern River induced by the latest Sierra uplift (Matthes, 1950a, p. 9—13). Some of the smaller side valleys, like those in the interval between Rock Creek and Golden Trout Creek, remained unglaciated and had their hang- ing aspect greatly enhanced by the glacial deepening and widening of the main canyon. The larger side valleys contained glaciers of their own, and therefore were themselves deepened and widened into U-shaped troughs, but nevertheless they remained hanging be- cause of the superior eroding power of the trunk glacier. (The small valleys on the Chagoopa Plateau were glaci- ated but on the whole probably underwent little change—that is, with reference to their gradients.) As a result of postglacial stream cutting, the lower reaches of the hanging valleys are now deeply trenched for some distance back from the Kern Canyon. The larger and more powerful streams, such as Wallace Creek, have made the most progress in cutting their hanging valleys down to the level of the master stream. The pronounced U-shaped form of the Kern Canyon has been evolved by glacial erosion from a narrow V- A54 shaped trench which the Kern River had cut as a result of the latest Sierra uplift. The canyon has been gla- ciated three times and is therefore a product of alter- nating stream and glacial erosion. Its present U-shaped form is not precisely the one which was cut by the gla- cier, for many changes have taken place in the canyon since glacial action ceased. The rock floor is. buried under thick deposits of boulders, gravel, and sand—~in part morainal, in part stream borne. The low gradient of the canyon floor over a stretch of several miles sug- gests that the glaciated rock floor underneath was ex— cavated into a chain of lake basins which are now filled and covered up. The walls, once smooth, are now fur— rowed by gullies; and talus slopes at the base of the walls form the curves of a new U—shaped form super- imposed upon the glacially eroded one. Deposition in the canyon is at present proceeding faster than erosion by the river. LOCAL GLACIATION SOUTH OF SEQUOIA NATIONAL PARK (WEST OF MAIN SIERRA CREST) SOURCES OF THE LITTLE KERN RIVER The extreme head of Shotgun Creek in Wisconsin time held a small glacier that excavated the cirque now occupied by Silver Lake. In the main valley of the Little Kern, above the junction with Shotgun Creek, Wisconsin moraines plaster the slopes on both sides of the Stream. These moraines were noted by Lawson (1904, p. 355), who wrote, “* * * huge lateral moraines were deposited notably on the east side of the canyon above the mouth of Shotgun Creek. This moraine has a height of perhaps 500 feet above the Little Kern, and spilled over the crest of the spur which separates Shotgun Creek from the Little Kern into the basin of the former.” There is here the record of a glacier of considerable volume that descended the valley from the vicinity of Vandever Mountain. Its length was about 6 miles, and it reached down to an altitude of 7,300 feet. This glacier was joined by another ice stream that cascaded down from the large cirque above Wet Mea- dows. The trail from lower Wet Meadows to Quinn Horse Camp crosses the lateral moraines of this system. Moraines of the older stage are probably present in Shotgun and Little Kern valleys, but opportunity to search for them was not provided. Glaciers also formed in the two valleys of Soda Spring Creek, immediately south of Wet Meadows. In the more southerly of these branches, vigorous gla- cial erosion has attenuated the ridge at the valley head, and the floor of the cirque is roughened by many con- verging bouldery moraines of both stages. Quinn’s Ranger Station is at the lower edge of a wet meadow held in by one of the many older moraines. The posi— SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY tions and outlines of the small glaciers that occupied the several headwaters of Soda Spring Creek in the Wisconsin Stage are clearly indicated by other moraines. At the head of Peeks Canyon is another broad com- pound cirque, in which lie a number of scattered tarns. This cirque—the rim has an altitude of 9,800 feet— Was produced by a shallow glacier that, at least in Wis- consin time, lay on the slopes southeast of the ridge bearing Sheep Mountain. SOURCES OF THE TULE RIVER On the northwest side of Sheep Mountain, at one of the headwaters of the Tule River basin, is another tarn, Summit Lake, in a little cirque that straddles the south boundary of Sequoia National Park. The lake is held in partly by ledges of metamorphic rock and partly by a small moraine. The moraine has been supplemented by an artificial dam, 4 to 5 feet high, that formerly had a gate with a handwheel and screw stem to raise it. These no longer function properly, and the water finds its way through the dam as best it can. Northwest of Summit Lake, two other valley heads, also draining into tributaries of the Tule River, have been mildly glaciated, the westerly one apparently only in the earlier stage. Headwalls of the cirques here and at Summit Lake are only 9,000 to 9,500 feet in altitude. SOURCES OF GOLDEN TROUT CREEK AND THE SOUTH FORK OF THE KERN RIVER Southeast of Sequoia National Park, five glaciers formed on the southerly flanks of the ridge between Rock Creek and Golden Trout Creek basins. These glaciers originated in cirques whose headwalls are at altitudes of 11,000 to 12,000 feet, and they descended the tributary valleys of Golden Trout Creek distances of 1 to 31/2 miles in the Wisconsin Stage and somewhat great- er distances in El Portal Stage. During the earlier gla- ciation, the three glaciers south of Siberian Pass united to form a common ice mass that spread widely over Whitney Meadows. East of this rank of glaciers, a sixth and even smaller ice body formed on the west side of the main Sierra crest, about a mile southeast of Cirque Peak. A typical cirque, containing a lake, attests to the existence of this glacier. Thirteen miles farther southeast on the Sierra crest, at one of the headwaters of the South Fork of the Kern River, a small glacier formed on Olancha Peak. Situated at an altitude of about 10,000 feet at lat 36°15’ N., this glacier (not shown on pl. 1) may have originated farther south than any other in the Sierra Nevada. South of the Golden Trout Creek basin, the configura- tion of certain valley heads on the steep northerly slopes of the Toowa Range (highest summit, Kern Peak, 11,- GLACLAL RECONNAISSANCE OF SEQUOIA NATIONAL PARK 493 ft) indicate that the range bore several small gla- ciers, at least in the later glacial stage. The most east- erly of these glaciers discharged into one of the sources of the South Fork of the Kern River. SELECTED REFERENCES Alden, W. C., 1932, Physiography and glacial geology of eastern Montana and adjacent areas: US. Geol. Survey Prof. Paper 174, 133 p. Axelrod, D. I. and Ting, W. S., 1960, Late Pliocene floras east of the Sierra Nevada: California Univ. Dept. Geol. Sci. Bull., v. 39, no. 1, p. 1—118. Birman, J. H., 1954, Pleistocene glaciation in the upper San Joaquin basin, Sierra Nevada, [pt] 6, in Jahns, R. H., ed., Geology of southern California: California Div. of Mines Bull. 170, p. 41—44. 1957, Glacial geology of the upper San Joaquin drainage, Sierra Nevada, California: Los Angeles, California Univ., Ph. D. thesis, 237 p. (unpublished). Blackwelder, Eliot, 1931, Pleistocene glaciation in the Sierra Nevada and Basin Ranges: Geol. Soc. America Bull., v. 42, no. 4, p. 865—922. Brewer, W. H., (Farquhar, F. P., ed.), 1930, Up and down Cali— fornia in 1860—1864: New Haven, Conn., Yale Univ. Press, 601 p. Durrell, Cordell, 1940, Metamorphism in the southern Sierra Nevada northeast of Visalia, California: California Univ. Dept. Geol. Bull., v. 25, no. 1, p. 1—117.‘ Farquhar, F. P., 1926, Place names of the High Sierra: San Francisco, The Sierra Club, 125 p. 1929, The story of Mount Whitney: Sierra Club Bull., v. 14, p. 39-57. Flint, R. F., 1957, Glacial and Pleistocene geology: New York, John Wiley & Sons, 553 p. Flint, R. F., and others, 1945, Glacier map of North America: Geol. Soc. America Spec. Paper 60, 37 p. Fry, Walter, and White, J. R., 1930, The Big Trees: Palo Alto, Calif, Stanford Univ. Press, 126 p. Fryxell, Fritiof, 1956, Memorial to Francois Emile Matthes (1874—1948) : Geol. Soc. America Proc. 1955, p. 153—168. Fryxell, Fritiof, ed., 1962, Francois Matthes and the marks of time: San Francisco, The Sierra Club, 192 p. Gilbert, G. K., 1904, Systematic asymmetry of crest lines in the High Sierra of California: Jour. Geology, v. 12, p. 579—588; Sierra Club Bull., v. 5, p. 279—286 [1905]. Hake, B. F., 1928, Scarps of the southwestern Sierra Nevada, California: Geol. Soc. America Bull., v. 39, no. 4, p. 1017— 1030. Hills, T. M., 1928, Glaciation of the upper Kern Canyon and its tributaries: Sierra Club Bull., v. 12, p. 17—19. Hinds, N. E. A., 1952, Evolution of the California landscape: California Div. Mines Bull. 158, 240 p. Hudson, F. S., 1960, Post-Pliocene uplift of the Sierra Nevada, California: Geol. Soc. America Bull., v. 71, no. 11, p. 1547— 1573. Jenkins, 0. P., 1938, Geologic map of California: California Div. Mines, scale 1 : 500,000. J epson, W. L., 1910, The Silva of California : Berkeley, California ' Univ. Mem., v. 2, 480 p. King, Clarence, 1872, Mountaineering in the Sierra Nevada: Boston, James R. Osgood & 00., 292 p. A55 Knopf, Adolph, 1918, A geologic reconnaissance of the Inyo Range and the eastern slope of the southern Sierra Nevada, California, with a section. on The stratigraphy of the Inyo Range, by Edwin Kirk: U.S. Geol. Survey Prof. Paper 110, 130 p. Knopf, Adolph, and Thelen, P., 1905, Sketch of the geology of Mineral King, California: California Univ. Dept. Geology Bull., v. 4, no. 12, p. 227—262. Lawson, A. C., 1904, The Geomorphogeny of the upper Kern Basin: California Univ. Dept. Geology Bull., v. 3, no. 15, p. 291—376. 1906, The geomorphic features of the middle Kern [Ca1- ifornia]: California Univ. Dept. Geology Bull., v. 4, no. 16, p. 397—409. LeConte, J. N., 1907, The High Sierra of California: Alpina Americana, no. 1, p. 1—16. Matthes, F. E., 1926, Kings River Canyon and Yosemite Valley: Sierra Club Bull., v. 12, no. 3, p. 224—236. 1929, Multiple glaciation in the Sierra Nevada: Science, new series, v. 70, p. 75—76. 1930, Geologic history of the Yosemite Valley: U.S. Geol. Survey Prof. Paper 160, 137 p. 1933, Geography and geology of the Sierra Nevada, in Middle California and western Nevada: Internat. Geol. Cong, 16th, Washington, D.O., Guidebook 16, p. 26—40. 1937, The geologic history of Mount Whitney: Sierra Club Bull., v.22, no. 1, p. 1—18. —--—- 1938, Avalanche sculpture in the Sierra Nevada of Cal- ifornia: Internat. Geodetic and Geophys. Union, Assoc. Internat. Hydrol. Sci. Bull. 23, p. 631—637. 1947, A geologist’s view, in Peattie, Roderick, ed., The Sierra Nevada: New York, Vanguard Press, p. 166—214. Matthes, F. E. (Fryxell, Fritiof, ed.), 1950a, Sequoia National Park, a geological album: Berkeley, California Univ. Press, 136 p. 1950b, The incomparable valley, a geological interpreta- tion of the Yosemite: Berkeley, California Univ. Press, 160 p. Matthes, F. E., 1960, Reconnaissance of the geomorphology and glacial geology of the San Joaquin Basin, Sierra Nevada, California: U.S. Geol. Survey Prof. Paper 329, 62 p. Mayo, E. B., 1941, Deformation in the interval Mount Lyell— Mount Whitney, California: Geol. Soc. America Bull., v. 52, no. 7, p. 1001—1084. 1947, Structure plan of the southern Sierra Nevada : Geol. Soc. America Bull., v. 58, no. 6, p. 495—504. Miller, W. J ., 1931, Geologic sections across the southern Sierra Nevada of California: California Univ. Dept. Geol. Sci. Bull., v. 20, no. 9, p. 331—360. Muir, John, (Colby, W. E., ed.), 1950, Studies in the Sierra: San Francisco, The Sierra Club, 103 p. Russell, R. J., 1938, Climates of California: Berkeley, Cali- fornia Univ. Pubs. Geography, v. 2, p. 73—84. 1947 , Sierra climate, in Peattie, Roderick, ed., The Sierra Nevada—The range of light: New York, Vanguard Press, p. 323—340. Sierra Club, 1893—1961 : Sierra Club Bull., v. 1—46. Starr, W. A., Jr., 1953, Guide to the John Muir Trail and the High Sierra region: 5th ed., San Francisco, The Sierra Club, 130 p. Sudworth, George B., 1908, Forest trees of the Pacific slope: US. Dept. Agriculture Forest Service, 441 p. Voge, Hervey, 1937, The age of the top of Mount Whitney: Science, new series, v. 86, supp. 2226, p. 6. A56 Voge, Hervey, ed., 1954, A climber’s guide to the High Sierra: San Francisco, The Sierra Club, 301 p. Webb, R. W., 1936, Kern Canyon fault, southern Sierra Nevada: J our. Geology, v. 44, no. 5, p. 631—638. 1938, Relations between wall rock and intrusives in the crystalline complex of the southern Sierra Nevada of Cali— fornia: Jour. Geology, v. 46, no. 3, p. 310—320. 1946, Geomorphology of the middle Kern River Basin, southern Sierra Nevada, California: Geol. Soc. America Bull., v. 57, no. 4, p. 355—382. SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY Webb, R. W., 1950, Volcanic geology of Toowa Valley, southern Sierra Nevada, California: Geol. Soc. America Bull., v. 61, no. 4, p. 349—357. 1952, Upland meadows of southern Sierra Nevada, Cali- fornia [abs]; Geol. Soc. America Bull., v. 63, no. 12, p. 1309. White, J. R., and Pusateri, S. J., 1949, Sequoia and Kings Canyon National Parks: Palo Alto, Calif., Stanford Univ. Press, 212 p. Whitney, J. D., 1865, Report of progress and synopsis of the field work from 1860—1864, v. 1 of Geology: California Geol. Survey, 498 p. A Page Alabama Hills ................................ A10 Alluvial fans, Kern Canyon __________________ 15 Little Kern Lake ______________ 50 Alta Meadow _______________ 33 Alta Peak, altitude... ......... 6 asymmetric profile ........................ 33 Ash Peak, altitude ........................... 5 Aster Lake ................................... 29 B Benchlands, Kern Basin ...................... 7 Big Arroyo Canyon, glacial erosion ........... 39 Wisconsin Stage glaciation ................ 53 Big Arroyo glacier, Wisconsin Stage .......... 53 Bighorn Plateau .............................. 41, 51 Black Kaweah, altitude-_ 8 Black Kaweah Peak .......................... 8,39 Boreal Plateau, altitude ...................... 8 Boulders, El Portal glaciation .......... 22, 29, 47, 49 erratic, Bighorn Plateau .................. 51 Glacier Point deposits. Whitney glacier .......................... 43 Wisconsin moraines ...................... 20 Buck Canyon glaciation ...................... 28,33 C Cahoon Meadow .............. 1 .............. 28 Cascade Range, Sierra-Cascade chain.. _ 18 Castle Rocks, altitude ................. . 6 Chagoopa Falls ............................... 15 Chagoopa Plateau, altitude ................... 8 El Portal moraines .............. . 51 Glacier Point deposits. . 52 Wisconsin glaciation..- . 40 Cirque Peak, altitude ........................ 9 Cirques, Great Western Divide ......... 34, 38,39, 53 Kaweah Basin ............... ._.- 28,32 34 Kaweah Peaks Ridge ...... 39 Kern-Kaweah glaciation .................. 41 Kings-Kern Divide ....................... 53 Merced and ’l‘uolumne Basins 35 Mount Guyot ............................ 45 Pecks Canyon ............................ 54 Sierra Nevada main crest ...... 54 Summit Lake ................. 54 Tulainyo Lake .......... 42 upper Kern Basin ........................ 36 Wallace Lake ............................. 42 Wet Meadows- ...... 54 Clifl Creek ..................... 32 Clover Creek valley, El Portal Stage.- 29 Wisconsin Stage .......................... 28 Coast Ranges ................................. 18 Conterno Trail.. . 37, 46,47 Cordilleran ice ............. . 17 Coyote Creek, alluvial fan ..... . 16 junction with Kern Canyon ______________ 45 valley .................................... 38, 46 Coyote Peaks ............. 50 Crabtree Canyon, glaciation.. ......... 43 D Death Valley .................. 12 Deer Creek valleys ........................... 34 Dennison Ridge .............................. 35 Description of the area ........ . .............. 2 Dikes, aplite .................. ,. .............. 2 4 INDEX [Major references are in italic] E Page Eagle Lake vicinity .......................... A34 East Fork glaciation, Kaweah Basin .......... 34 El Portal Stage ................... 19, 22, 29, 36, 45, 53 Emerald Lake ................................ 28 Evelyn Lake ................................. 34 F Fault zone, Kern Canyon .................... 12 Fieldwork chronology ........................ 2 Florence Peak ............ 28 Fork Canyon, Kaweah Basin, altitude ....... 34 G Generals Highway, El Portal Stage ........... 29 Glacier Point Stage ....................... 30 Geographic sketch ............................ 2 Geology, evolution of landforms ..... . 2 Geomorphology, ancient glaciers ..... . 2 Giant Forest, altitude ........... 6 Glaciation, characteristics ................. 17,26,315 differentiation of stages ................... 19 geology ........................ . 2 Kern Canyon ................... . 37 south of Sequoia National Park-... 54 Glacier Point Stage ............... 19, 24, 30, 37, 52, 53 Golden Trout Creek .......................... 36, 54 Golden Trout Creek valley... . 37, 46, 47, 53 Granite, El Portal glaciation- - .. 49 vicinity of Wallace Lake ......... - 42 Grasshopper Flat, Glacier Point Stage ........ 52 Grasshopper Meadow ........................ 48, 50 Great Western Divide... 2, 4, 7, 28, 34, 36, 38, 39, 40, 53 _____________________ 45, 51 Guyot Flat ................................... 43 H Hamilton Lakes, Middle Fork Basin ......... 32 Heather Lake ................................ 28 High Sierra, glaciation ............... . 18 timber and lakes .................. .. 4 High Sierra Trail, Chagoopa Plateau.. 8, 33, 39 Hockett Lakes-Garfield Grove trail ........... 35 Hockett Meadows ............................ 34 Hole-in-the-ground, El Portal glaciation- 48 Homers Nose, altitude ....... 6 Horse Creek valley, glaciation ................ 34 I Ice Age, concept .............................. 17 Illinoian Stage ................................ 24 Inyo Range, east wall of Owens Valley ....... 10 I John Muir Trail .............................. 41, 42 Junction Meadow, head of Kern Canyon ..... 12 K Kansan Stage ............................. 24 Kaweah Basin, geographic sketch- ...- 6 Glaciatlon ................................ 26' East Fork ............................ 34 Marble Fork. . 98 Middle Fork-. 30 South Fork.._. 34 Kaweah Canyon .......................... 28, 29, 30 Kaweah Gap ................................. 39 Page Kaweah Peaks Ridge ........................ A39, 40 Kaweah River canyon, depth ................. 5 Kern Basin, upper, altitude .................. 36 Upper, geographic sketch . glaciation ..... Kern Canyon, depth. geomorphology ........................... 12 glaciation ................................. 36 Kern Canyon Ranger Station ............ 37 Kern glaciation, Glacier Point Stage, depth... 52 trunk glacier ........................ 15 Kern glacier system, members ................ 36 Kern-Kaweah glacier ......................... 40 Kern Lakes, origin. . Kern Peak, altitude- Kern Ridge ...... Kern River .......................... Kernbut ............................. Kerncol ...... Kings Canyon.. Kings Canyon National Park ................ 2 Kings-Kern Divide .............. Kings River glaciation ........................ L Lakes, Great Western Divide ................. 39 Laurel Creek canyon ............ 38 Laurel glacier .................... 46 Laurentide glaciation... _ 21 Little Kern Lake ............................. 50 Little Kern River ............................ 36, 64 Little Sand Meadow valley ............ 33 Lyell Canyon, Yosemite National Park ...... 39 M Marble Fork glaciation, Kaweah Basin ....... 28 McGee Stage ................................. 24 Mehrten Creek Valley ........................ 33 Merced Basin ..................... 35 Merced Canyon, El Portal moraines ......... 45 terminus of Yosemite glacier .............. 18 Merced River ..................... 42 Middle Fork, Kaweah Basin, glaciation ...... .90 Middle Fork Canyon, Kaweah Basin ........ 30 Middle Fork valley, Kaweah Basin .......... 32 Milk Ranch Peak, altitude ................... 5 Mineral King region, glaciation..- ...... 2 Mono Lake, glaciation ................... 19 Morame Lake, moraine 1mpounded. ._.. 39 Morames, El Portal Stage .................... £2 45 Glacier Point Stage ................... 24 identity of glacial stages. ...- 19 Kern Canyon..- ..- ..._ 14, 37 Wisconsin Stage. 20, 28,32, 34, 46, 54 Mount Barnard .............................. 42 Mount Hitchcock, avalanche chutes .......... 43 Mount Kaweah, altitude ........ 8 glaciation ............ 40 Mount Langley, its altitude and its similarily to Mount Whitney ............... 9 Mount Muir ....................... 43 Mount Russell ........ 42 Mount Tyndall .............................. 41 Mount Whitney, altitude .................... 43 general form ............... 9 Mount Young, avalanche chutes ............. 43 A57 A58 N Page Nehraskan Stage ............................. A26 Nevada Falls ................................. 42 Nunataks .................................... 35 0 Olancha Peak ________________________________ 54 Owens Valley ________________________________ 10 P Panther Peak, altitude ....................... 7 Paradise Peak, altitude ....................... 6 Pear Lake ___________ 28 Pecks Canyon _______________________________ 54 Pleistocene Epoch, glaciation _________________ 19 Precipitation. Great Western Divide.-.._ . . . . 36 Q Quinn Horse Camp .......................... 54 R Rattlesnake Creek canyon .................... 38 Rattlesnake glacier ........................... 53 Red Kaweah Peak . 8,39 Red Spur ..................................... 8, 40 Red Spur falls ................................ 15 Redwood Meadow. 32 Rock Creek __________________________________ 53 Rock Creek basin ____________________________ 54 Rock glacier _____________ 41, 45, 52 Rockslides, Kern Canyon .................... 50 S Sand Meadows ............................... 35 Sandy Meadow .............................. 43 Sevenmile Hill trail- 33 Sheep Mountain ............................. 54 Sherwin Stage ................................ 24 INDEX Page Shotgun Creek ............................... A54 Sierra Nevada, limits of glaciation ____________ 18 main crest ....................... 10, 36, 41, 53, 54 Sequoia National Park sharp-profiled peaks ...................... 9 Sierra-Cascade chain _____________________ 18 southernmost point 0! glaciation._ 53 Silliman Creek valley, El Portal Stage ________ 29 Wisconsin Stage .......................... 28 Silllman crest, altitude. 28 Silver Lake ___________________________________ 54 Sky Parlor Meadow .......................... 39, 51 Soda Spring Creek valleys... _ 54 South Fork, Kaweah Basin. glaciation ________ 84 South Fork Canyon, Kaweah Basin, altitude. 34 South Fork Kern River ...................... 36, 54 South Fork Rock Creek ...................... 45 Summary statement, Kern Canyon glaciers.. 53 Summit Lake ................................ 54 T Table Mountain, altitude .................... 9 Tableland, altitude ........................... 28 Tahoe glaciation.._. 21 Takopah Valley .............................. 28 Talus slopes, Kern Canyon ................... 54 Tarns, Kaweah Basin.-_. 28 Peeks Canyon ............................ 54 Summit Lake ............................ 54 Terraces, Clifl‘ Creek valley-- 32 Timber Gap Creek valley .................... 32 Tioga glaciation .............................. 21 'l‘oowa Range. . 54 Tower Rock .................................. 53 Triple Divide Peak .......................... 28 Trout Meadows ........ 48, 49 Tulainyo Lake, altitude ...................... 42 Tule River ................................... 36, 64 O Page Tuohy Meadow, glacial boulders ............. A34 Tuolumne glaciation ............... 18,35 Tyndall Creek ..................... 51 Tyndall Creek canyon. 38 Tyndall glacier ....................... . 41 U Upper Kern Basin, altitude .................. 36 geographic sketch _____ 7 glaciation ..................... 18, 36 Upper Whitney Canyon ...................... 43 V Vandever Mountain... 54 Vernal Fall ................................... 42 W Wales Lakes ............ Wallace Creek Wallace glacier ......... Wallace Lake, granite. . Waterfalls Kern Canyon. Wet Meadows ................ 54 White Mountains, Inyo Range. 12 Whitman Creek valley ...... 34 Whitney Creek _______ 51 Whitney glacier- _ 41, 43 Whitney Meadows. ._ . 54 Wisconsin Stage--.. Wright Creek. - - _ - - . - - 51 Wright glacier.. .................. 41 Y Yosemite glacier ........... 18 Yosemite National Park ...................... 2 Yosemite Valley, El Portal moraines in Kern Canyon .......................... 45 Wisconsin moraines in Kern Canyon ..... 37 UNITED STATES DEPARTMENT OF THE INTERIOR GEOLOGICAL SURVEY 118°50’ 36°45' , i" 2:111» 3’ u' I 51’ .v , 1’ 4; if" 7’ :f/ '1. , ‘j‘y; ' I I355 \ 6/! / Areas coyéfédj by ice of I the ' Wisconsin Giacial Stage ' . “ ' Areas covered by ice Of the El Portal Glacial Stage Location of moraine of the Glacier Point Glacial Stage 5/.\\‘\‘f‘ ii'fi“ Distinct crests of Wisconsin mo- raines in the Vicinity of the Kern Canyon Ranger Station Boundary of area covered by ice Dashed where approximate 36°16’ ' 118°50’ Base from US. Geoiogical Survey topographic map: Sequoia and Kings Canyon National Parks, California, 1937 16" I L Cr 0 2 o 1: Lu z U v TRUE NORTH E APPROXIMATE MEAN DECLINATION,1965 . I 4‘ I ,, Kiri/Ag .5241 ii ' @Iiii‘z‘ 1;,”in ri/I‘I’w " " ‘fll’l'cfih‘ ’ 2/ F b , ! 4 ii at, iii/I N . , In ' 't f , \ i an. ,; WM 1117’ ” '5' I'll ’I l i, ’5 I I ”(ml/l w?! .4 M” gzll‘li'lzligu 'r l- I”? ’i” "1/1 I kit/5. ‘! I a I; if?“ i y ’ ‘tg’IIf/iw’ if? 11"}???- J’W " "’ 4 . w flu, viii!" ."‘ II." v 2* 4; I It ”In " ‘ A ///// 'f'i,‘ .v’i' "i p : ; , {iii/.5 «'1 4%, MAP OF ANCIENT GLACIERS OF SEQUOIA 5-: 6‘: Iii"? é .V f “II: = is w - i! i i In,» 30' ”ii: Wilt/II g, ' 1’7" ,g e-‘ I"!!! 37$!le 1"; (.4 1‘ 9 i/ a” . I if". «7'! M: ! iiéi’ if: 3‘ ' Ju "I ’9" 4%.!!! 1,1 : : ~}fi;% I! 1’, iii/«:1 3W 1 I aéf'f’” w '1; ‘ > ,1! iiw'k‘ I 11L” iii}: " //: "" W 3!th if} I 0"”. ~» 'I‘IZI’II‘i‘III! All! ”I! Iii“: i W: i . ‘i f? (iii/III ii ‘ fix?" 45””?! 7112’}, a 9 A! i 1': _ 20' i _ 35;? J6” I ., i ii .. i4 ‘ nil/If“?! ”gm; 71%” a g. I. 5 I}, [if]? [if 5 1w // ; 554154555 2%? PLATE 1 PROFESSIONAL PAPER 504—A 118° 10’ - 36°45’ it! .“ .41} i /. .1. 1,, WWW/W 30’ i 20! SCALE 1:125 000 NATIONAL PARK, SIERRA NEVADA, CALIFORNIA 10 MILES i——-———-————”I 4 6 2 4 6 8 I—‘———I I—-—~—"*‘I I——-fi CONTOUR INTERVAL lOO FEET DATUM IS MEAN SEA LEVEL 10 KILOMETERS I / L a v .‘ ,3” ,, my ., IOR—GEOLOGICAL SURVEY. WASHINGTON, D C 118° 10" Mapped by Francois E. Matthes 36°16" UNITED STATES DEPARTMENT'OF THE INTERIOR I PROFESSIONAL PAPER 504—A GEOLOGICAL SURVEY . PLATE 2 g < "S o '5 § C LIJ L Z 7 ,' 3 < ._ -- a 5 A $ A? g E : 3 I3 (0 14,000’ , /\ 14,000' - TE 1’ \ Iv 53 (D C \\\‘fi I v: 0 /r\ \\ _ _ § >q / \ \ s g //, \ 7 g \\\\ O ///F. \\J \\\ I , M \\\\Hifi C / \ -10000 10,000 — a \\ L3\\ 5 / \\ » ‘ :5 \\\_\_\\\ K / \ 7 V ‘ S / \\ » 3 \ - c \’/ \ _ ”3 \ I \ , 5000' ‘ ,7 ’ 7 5000' ' VERTICAL EXAGGERATION X2 < < 2 o > < N i Z z < < E E K c 5 § 5 g. 0‘) >‘ “I C a 14,000' g 9 14,000' 14,000' § g 14,000, 0 § 52 g L C S 0 10,000' “\\\ 3 9 10,000' 10,000' m C 10,000, L \‘x// g \“\ Q , O \/ 5000, - , 5000’ 5000' 5000' NO VERTICAL EXAGGERATION NO VERTICAL EXAGGERATION PROFILES THROUGH SEQUOIA NATIONAL PARK, CALIFORNIA SCALE 1:125 000 if, 710 MILES igo KILOMETERS 748-960 0 — 65 (In pocket) Q g 175/ 7 DAY 7% #5’M~6’ . . Postglaaal Dramage Evolution and Stream Geometry in the Ontonagon Area, Michigan GEOLOGICAL SURVEY PROFESSIONAL PAPER 504—B fig” 0F CAL/f0\\ (k 00m; 1965 y" ">571 Q5 p v \WA/ Postglaeial Drainage Evolution and Stream Geometry in the Ontonagon Area, Michigan By JOHN T. HACK SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY GEOLOGICAL SURVEY PROFESSIONAL PAPER SO4—B A, para/lei drainage networé contra/[ed a}! glacial graows provider a'az‘a f0r z‘fle staa’y of fluvial geomorp/zic p/zeaomma, expecially t/ze geometry 0fsz‘ream égflfl’l'dl‘lbflf UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON :1965 UNITED STATES DEPARTMENT OF THE INTERIOR STEWART Lt UDALL, Secretary GEOLOGICAL SURVEY Thomas B. Nolan, Director For Salé by theJSuperrintendent ovaocumehfs;V US. Government Printihg Officg ' ’ ' Wash-ingtOn, DC. 20402 ' ‘ C O N T E N T S Page Evolution of the stream valleys—Continued Page Abstract ........................................... Bl Size of bed material _____________________________ B20 Introduction _______________________________________ 1 Channel cross section ____________________________ 21 Climate and streamflow _____________________________ 2 Cranberry River ____________________________________ 22 ISDIIJmmary ogbedrcgck geology """"""""""""" : Terraces of the lower Cranberry River _________________ 27 eistocene e osi s _________________________________ , Thicknesspof the glacial drift _____________________ 5 Potato River """"""""""""""""""""""" :3 Recent deposits _____________________________________ 5 Meanders ------------------------------ - n 29 Surface features of the Ontonagon Plain _______________ 5 CODdltlons that prodlice meanders ---------------- Glacial grooves or flutes _________________________ 5 Meander wavelength In the Ontonagon area ________ 30 Glacial lake shorelines ___________________________ 7 Meander wavelengths in the Shenandoah Valley of Age of the shorelines ________________________ 7 Virginia ------------------------------------- 34 Kettle holes ____________________________________ 8 Conditions controlling meander wavelength ________ 34 Beaver ponds __________________________________ 8 Summary and conclusions ____________________________ 36 Chronology of emergence of the Ontonagon Plain _______ 9 The drainage system ____________________________ 36 Evolution of the stream valleys- ______________________ 9 Evolution of the landscape _______________________ 37 Drainage composition ___________________________ 10 Graded streams ................................. 39 Longitudinal profiles ____________________________ 17 References cited ____________________________________ 39 ILLUSTRATIONS Page PLATE 1. Map of the Ontonagon area, Michigan _____________________________________________________________ In pocket FIGURE 1. Relation of discharge to drainage area in streams of northern Michigan ____________________________________ 32 2- Map Of geologic and geographic subdivisions of Ontonagon area __________________________________________ 3 3. Minor surface features of the Ontonagon area __________________________________________________________ 6 4. Relation of valley depth to drainage area ______________________________________________________________ 10 5. Profiles of small streams _____________________________________________________________________________ 11 6. Examples of drainage systems ________________________________________________________________________ 12 7- Relation 0f stream order to stream length and number of streams _________________________________________ 13 8. Possible relations of stream length to area under various conditions _______________________________________ 15 9- Relation of length to drainage area of streams on the Ontonagon Plain ____________________________________ 16 10- Model drainage network simulating streams of Ontonagon Plain __________________________________________ 17 11. Relation of channel slope to stream length _____________________________________________________________ 19 12. Longitudinal profiles of three streams _________________________________________________________________ 20 13. Relation of channel slope to drainage area ______________________________________________________________ 21 14- Relation 0f Channel slope to ratio of bed-material size to drainage area _____________________________________ 21 15- Relation 0f Channel Width and depth-width ratio to drainage area _________________________________________ 22 16. Map of lower reaches of Cranberry River____________________________________________________________;_ 23 17. Views of Cranberry and Potato Rivers ________________________________________________________________ 25 18. Profile of lower valley of the Cranberry River __________________________________________________________ 28 19. Lower reaches of the Potato River ____________________________________________________________________ 29 20- Relation of channel slope to drainage area in meandering and nonmeandering streams _______________________ 31 21- Typical meandering reEChes of the Ontonagon area _____________________________________________________ 32 22. Relation of meander wavelengths to drainage-basin areas _________________________________________________ 33 23- Detailed plan of short reach showing third-order bends __________________________________________________ 33 24- Relation of meander wavelength to drainage area in the Shenandoah Valley, Va ............................ 34 25- Comparison of typical channel cross sections in alluvial and bedrock reaches ________________________________ 35 TABLES Page TABLE 1. Characteristics of the drainage net of certain stream basins and groups of basins in the Ontonagon area _______ 313 2- Measurements in stream valleys at localities in the Ontonagon area _______________________________________ 19 3. Comparison of features of the channel of the Potato River in the bedrock canyon with those of the till area down— 28 stream __________________________________________________________________________________________ III SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY POSTGLACIAL DRAINAGE EVOLUTION AND STREAM GEOMETRY IN THE ONTONAGON AREA, MICHIGAN By JOHN T. HACK ABSTRACT The clay plain on the south shore of Lake Superior in Ontona- gon County, Mich., is underlain by glacial deposits of Valders and post-Valders age consisting of till interbedded with fine- grained glaciolacustrine sediments. The surface of the plain is strongly grooved by glacial flutings parallel to the direction of ice motion. The grooves have a regular wavelength that averages 380 feet. In places the grooves are buried by laws— trine sediments of glacial Lake Duluth. Seven or more glacial lake shorelines cross the plain, in places interrupting the grooves. Six shorelines between 650 and 1,200 feet above sea level are attributed to glacial Lake Duluth. One ancient shore- line at about 615 feet altitude is attributed to Lake Nipissing. Lake Duluth was short lived. It formed and withdrew com- pletely between 10,300 and 9,500 before present. More con- spicuous shore features were formed by Lake Nipissing at its stand only 15 feet above the level of present Lake Superior. As Lake Duluth withdrew from the Ontonagon Plain, a drain- age network formed. The streams followed the grooves that extended downslope toward the lake. The resulting drainage pattern is strongly attenuated downslope, but, except for the attenuation, is similar in its geometry to drainage patterns in well-graded landscapes. The evidence available supports the theory that the spacing of stream bifurcations or junctions in a typical drainage network is random and that stream lengths tend to be proportional to the six-tenths power of the drainage area because of the random arrangement of channels. Very little change has taken place in the Ontonagon drainage pattern as a result of piracies or headward migration of divides. The depth of incision of the present valleys is roughly proportional to a function of the drainage area and there is little or no evi- dence to support the idea of nickpoints that migrate upstream. Instead, the valleys are cut to depths proportional to- the avail- able discharge. The longitudinal profiles of the larger streams are slightly concave. The streams are adjusted to the bed and bank mate- rials and those in bedrock have steeper gradients than those in unconsolidated deposits. Most of the streams have intricate meander patterns, both of their channels and their valleys. Streams flowing in bedrock, however, have straighter courses. Meander wavelength increases in a regular manner as average discharge increases, and is also affected by the materials in the valley; streams in bedrock have meanders about four times the wavelength of streams in till. It is concluded that the evolution of the drainage system in the Ontonagon area has proceeded in response to upstream factors, such as available discharge and the geology of the drainage basin. Changes in base level have affected the streams mainly by changing the relief and the available potential energy in the landscape. INTRODUCTION The Ontonagon area includes an extensive clayey plain south of the Lake Superior shore. This plain, referred to herein as the Ontonagon Plain, is deeply trenched by an unusual drainage network that has de— veloped since the Valders glaciation and the draining of glacial Lake Duluth about 9,500 years ago. Because of the unusual nature of the drainage network, the short time in which it has developed, and the controls available by which to date the development, the area is one where much can be learned about the evolution of a drainage system and some of the geologic factors that are involved in the equilibrium conditions in stream channels. The plain lies on the south shore of Lake Superior between the Porcupine Mountains and the F iresteel River; it is bordered on the south by a range of bedrock hills, known as the Copper Range, that rises more than 1,000 feet above the lake. The plain is about 8 miles wide and slopes toward the lake with an average gra— dient of 50 feet per mile. It is underlain by fine- grained glacial-lake sediments and till. A remarkable series of long shallow grooves crosses the plain more or less parallel to the direction of slope. Because the grooves have controlled drainage development, the streams flow from the Copper Range across the plain in straight but narrow valleys and have narrow atten- uated drainage basins. Most of the study of the drainage network was done by means of aerial photographs and topographic map interpretation. Photogrammetric techniques were em- ployed in studies of some areas. A total of 9 weeks, however, in 1958 and 1960 was spent in the field in order to examine the materials that enclose the stream chan- nels and to make observations of the bed materials of B1 B2 the streams at various localities (pl. 1). Because the deposits beneath the plain have not previously been described, a reconnaissance study was made of the glacial geology. The stratigraphy and petrography of the glacial deposits beneath the Ontonagon Plain were studied by Friedrich Brandtner who was in the field with the writer in 1960. The choice of the area as a place to study a drainage network was suggested by Walter S. White of the US. Geological Survey, who noted the grooves and recog- nized that they were produced by glacial scouring. Much of the detailed study of glacial deposits was done on the property of the White Pine Copper Co. Many courtesies and valuable help were extended by the company. ‘ CLIMATE AND STREAMFLOW The Ontonagon plain has a humid but cold conti- nental climate. At Bergland, south of the Copper Range, the mean monthly temperature ranges from 11°F in January to 67 .7°F in July. The average grow- ing season is only 84 days and precipitation in the form of snow is common from November to May. Precipita- tion averages 32 inches and is only slightly seasonal, be- ing heaviest in the late spring and summer and lightest in the fall. Snowfall just south of the Lake Superior shore averages 115 to 130 inches (U.S. Dept. Agricul— ture, 1941). Lake Superior has a marked ameliorating effect on the climate of the Ontonagon Plain, as is well shown by the vegetation. Whereas spruce and fir forests are prevalent south of the Copper Range, the area be— tween these hills and Lake Superior is occupied by a typical northern hardwood forest. The most abundant trees are hemlock, maple, birch, and aspen; oaks are found in limited numbers on dry, generally south-facing slopes. Within the study area, the only stream gage measur— ing flow uncontrolled by a dam is on the Ontonagon River west of Rockland (pl. 1). The gaged flow at this point averages 1,320 cfs (cubic feet per second) from a drainage area of 1,290 square miles (U.S. Geol. Survey, 1961). Examination of the records at other gaging stations on streams in northern Michigan trib- utary to Lake Superior indicates that the average dis- charge in cubic feet per second of each of the streams is roughly equal to its drainage area in square miles (fig. 1). Seasonal fluctuations are large. The highest aver- age flow over an extended period generally is in April and May when mean monthly discharges more than 5 times the mean annual discharge are common. Momen- tary peak discharges of more than 15 times the mean annual have been recorded in almost every stream, though none of the stations have a very long period of SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY 1400 ._. N 8 | \ I 1000 / 800 L 7 . / 600 ~ / 400 l— o / AVERAGE DISCHARGE, Q, IN CUBIC FEET PER SECOND m 8 I . O \ \ | | l | L 400 600 800 1000 1200 1400 DRAINAGE AREA, A, IN SQUARE MILES 0 I O 200 FIGURE 1.—Rela‘tion of average discharge to drainage area at gaging stations in northern Michigan west of the Sturgeon River, in streams tributary to Lake Superior. record. The peak flows are not necessarily in the spring. The largest flood in the region during the 18 years prior to 1960 occurred on August 18, 1942, when the discharge on the Ontonagon River at Rockland was estimated at 42,000 cfs, almost 40 times the mean. Comparison of the streamflow characteristics of the Ontonagon Plain with those of the central Appala- chians, where the writer has made similar studies (Hack, 1957), indicates that except for peak flows the stream regimens are not very different. On the Ontonagon Plain the seasonal variation in flow is more pronounced and regular. The peak recorded discharge, hOWever, is much less. In the Potomac River basin of Virginia, West Virginia, and Maryland, for example, many streams have recorded momentary discharges more than 100 times the mean annual. SUMMARY OF BEDROCK GEOLOGY The Ontonagon area is mostly within the outcrop area of the Keweenawan Series of Precambrian age. The character of the bedrock has a strong influence on the topography and on the material in the glacial de- posits. The following description of the rocks is taken from papers by Van Hise and Leith (1911) , Butler and Burbank (1929), White and Wright (1954), Hamblin (1961), and Hamblin and Homer (1961). Portage Lake Lewd Samba—These rocks of middle Keweenawan age underlie the hills known as the Copper Range (fig. 2). They are bordered on the south 'by the POSTGLACIAL DRAINAGE, STREAM GEOMETRY, ONTONAGON AREA, MICHIGAN Keweenaw fault and on the north are overlain uncon- formably by the Copper Harbor Conglomerate. The Portage Lake Lava Series includes more than 20,000 feet of interbedded lava flows and thin conglomerate. Most of the lavas are basaltic in composition but felsite and other silicic rocks are included in the lava series. In the Porcupine Mountains there is a large area of such rocks. The lava series contains many copper-bearing beds, particularly the interbedded conglomerate and scoriaceous tops of the lava flows. Once-famous mining districts no longer in production are at Norwich, Vic- toria, Rockland, Greenland, and Mass. Copper Harbor 00nglomemte.—The Copper Harbor Conglomerate of late Keweenawan age underlies the ’ 89°30’ Area shown on topographic POR PINE MOU TAINS 46°45’ O B3 northern and eastern slopes of the Porcupine Mountains and the north side of the Copper Range. The rocks con-‘ . sist mostly of reddish, rudely stratified conglomerate With rhyolitic boulders predominating. Medium- to coarse-grained sandstone, some thin and persistent, some lenticular, is interbedded in the conglomerate. Thin beds of volcanics also occur. The thickness of these rocks in the Ontonagon area is 2,300 to 5,500 feet (White and Wright, 1954) . Nonesuch Shale.——The Copper Harbor Conglomerate is overlain by a band of dark-gray siltstone, shale, and sandstone, generally 600 feet thick. These rocks (the Nonesuch Shale of late Keweenawan age) contain a cupriferous zone near the base which, at White Pine, is 89°15’ 89°00’ 5 10 MILES |_I_I_J__J__I—___I EXPLANATION PRECAMBRIAN OR PRECAMBRIAN ROCKS C‘AMBRIAN ROCKS ' Freda Sandstone and Nonesuch Shale J acobsville Sandstone Major fault Copper Harbor Conglomerate and Portage Lake Lava Series Contact FIGURE 2.—Simplified geologic map of the Ontonagon area showing major geographic and geologic divisions. B4 mined for copper. The Nonesuch Shale forms a band along the south edge of the Ontonagon Plain and under- lies an extensive area in the Iron River basin. It crops out only in the bottom of the deeper valleys and along the Lake Superior shore near Silver City (pl. 1). Freda Sandstone.—~This formation of late Kewee- nawan age consists of alternating layers of reddish— brown fine arkosic sandstone and red micaceous silty shale. Some conglomerate occurs near the base. The material is derived mostly from basaltic lavas. Quartz is not abundant. Ripple marks, raindrop imprints, and cross-stratification are common. The Freda Sand- stone is more than 14,000 feet thick and crops out ex~ tensively along the Lake Superior shore and in many valley bottoms of the Ontonagon Plain. Inasmuch as this sandstone underlies glacial deposits over a wide area and also, from the shoreline to well out into the lake, forms the lake bottom, it is undoubtedly the source material of much of the till that underlies the Ontonagon Plain, and has imparted to the till its char- acteristic reddish-brown color. Jacobsville Sandstone—The area south of the Copper Range is underlain by gently dipping sandstone beds of Precambrian or Cambrian age. The Jacobsville Sandstone is separated from the Portage Lake Lava Series by the Keweenaw fault. The rocks are quartz- itic rather than arkosic and, though reddish, are of considerably lighter shade than the Freda Sandstone. They are medium grained, well sorted, and crossbedded, and contain lenses of conglomerate. Outcrops in the Ontonagon area are not extensive and are confined to the valley of the Ontonagon River and its tributaries above Rockland. Structure—The structure of the Keweenawan rocks, except those in the Porcupine Mountains, is monoclinal, the dip being to the north. In the Copper Range, dips may exceed 50°, but they flatten to the north and, under the Ontonagon Plain, are generally less than 10°. The Porcupine Mountains are a structural dome or arch and have steep dips on both the north and south flanks. PLEISTOCENE DEPOSITS The glacial geology of the Lake Superior region is known principally from the reconnaissance work of Leverett (1928), though his chronology and nomencla- ture have been modified by more recent work in various parts of the region. This newer work has been sum- marized by Hough (1958, 1963). The Ontonagon area is entirely in the belt of late Wisconsin drift commonly referred to as Valders. The terminal moraine of the Valders Stade lies about 60 miles south of Ontonagon (Flint and others, 1959). A minor recessional m0— raine traverses the Copper Range, though till is thin on SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY most of the hills and bedrock exposures are numerous. The Ontonagon Plain and the Iron River basin are un— derlain by thick layered deposits of till and lake sediments. Till and outwash of the Copper Range and Porcu- pine Mountain8.—In the hilly areas of the region, drift is thin and outcrops are numerous, especially on the south sides of the hills. Most of this thin till is dif- ferent from the thick till of the Ontonagon Plain; it is more stony, contains volcanic rocks derived locally, and has a loamy matrix. One area of glacial outwash along Sleepy Creek in sec. 7, R. 42 W, T 49 N occupies about 1 square mile. Glacial till and lamtrine deposits of Ontonagon Plain and Iron River Basin—Most of the area north of the Copper Range is underlain by reddish—brown glacial till and lake deposits which are interbedded in many places. The arrangement of the deposits sug- gests that they were laid down by an ice sheet that re- treated into an ice marginal lake and then readvanced. In the western part of the area where they were ex- . amined in greatest detail, the deposits can be subdivided into three distinct beds referred to as the lower, inter— mediate, and upper units. The lower unit consists of stony till containing subangular boulders and frag- ments derived mostly from the bedrock under the On- tonagon Plain. The bedrock over most of the area is the Freda Sandstone, which imparts a silty to fine sandy texture and reddish~brown color to the till matrix. The intermediate unit is in distinct layers or strata consist— ing of till and laminated silt and clay believed to be lacustrine sediments. Some of the layers contain large erratic boulders of gabbroic and basaltic rock, though in general, this unit is much less stony than the lower unit. The upper unit is a clayey till that is much less stony than either of the lower units. There are few erratics of large size, and a smaller proportion of the stones is derived from the Freda Sandstone. Y ounger lacustrine deposits.—The upper clayey till is overlain in places by strongly laminated clay, silt, and sand that vary widely in thickness. These are typical lacustrine deposits presumably laid down in late Pleistocene lakes marginal to the ice. South of the Copper Range in the eastern part of the area, where the three branches of the Ontonagon River come together, the lacustrine deposits make up almost the entire sec- tion above the bedrock and are more than 200 feet thick. This area was occupied by an ice-marginal lake known as Lake Ontonagon (Leverett, 1928, p. 57). The lake formed when the glacier stood on the south side of the Copper Range and blocked the Ontonagon River Basin. When the ice retreated north of the Copper Range, a large part of the present Lake Superior basin was un- POSTGLACIAL DRAINAGE, STREAM GEOMETRY, ONTONAGON AREA, MICHIGAN covered, and an ice-marginal lake formed that occupied half of the basin and had its outlet near Duluth, Minn. The shoreline of the lake, called Lake Duluth, is below an elevation of 1,200 feet in the Ontonagon area. The lake extended through the lowest gaps of the Copper Range and its south shore lay south of the area shown in plate 1. The north shore was presumably bounded by ice and lay somewhere north of the area. The deposits of glacial Lake Duluth are mostly a thin veneer and are somewhat patchy in distribution. East of the Ontona- gon River they are, in places, 20 feet thick or more; they form a thick fill in the valleys of the Flintsteel and Firesteel Rivers where they rest on till. In most of the area west of the Ontonagon River the lacustrine de- posits form a veneer less than 2 feet thick. In some places they are lacking entirely. THICKNESS OF THE GLACIAL DRIFT The bedrock surface beneath the drift is quite irreg- ular and the pre-Valders topography of the Ontonagon Plain must have had valleys at least 100 feet deep. As shown in holes drilled by the White Pine Copper 00., the drift near the Iron River is in places more than 150 feet thick; in the vicinity of the tailings pond, how- ever, it is only 20 to 30 feet thick and it continues to thin to the north to less than 10 feet. Along the Cranberry River the drift ranges in thickness from more than 80 feet to a minimum of 10 feet. Along the Ontonagon River valley the drift in places is at least 200 feet thick and the absence of bedrock exposures in the valley walls suggests that the Ontonagon River follows a preglacial course. The present drainage system is deeply intrenched; in some places the valley walls are mostly bedrock and in others mostly drift. Unfortunately, not enough data are available to show the character of the preglacial topography under most of the Ontonagon Plain. RECENT DEPOSITS Shore deposits older them Lake NipissMg.—Sandy deposits, that are apparently partly dune sand, but that may in part be delta sand occur on the Ontonagon Plain near the Lake Superior shore. These deposits veneer the upland plain inland from the scarp that marks the shoreline of Lake Nipissing. They form a belt averag- ing about a mile wide that widens to 2 miles near the Ontonagon River. In part, they may be dunes blown up over the low escarpment from the Nipissing beach. Near the Ontonagon River, however, the sand is too thick to have had this origin, for it forms the entire escarpment as though deposited in a delta when the Ontonagon River was at a higher level. 748—958 0—65—2 B57 Lake Nipissi/ng and younger beaches.—~The Ontonagon Plain ends abruptly about 1,000 feet south of the Lake Superior shore where it drops from 5 to 30 feet to a low sandy belt that borders the lake. This belt is a complex of beach and dune sand, partly shore deposits of Lake Nipissing, which stood at an altitude of about 615 feet, and partly shore deposits of modern Lake Superior. The deposits are coarse sand and in places, especially marginal to the larger streams, gravel. Recent stmam deposits.—Most of the streams longer than a mile or two are intrenched in the till or in bed- rock, and are bordered by a floodplain or, in places, by terraces. These deposits are generally reddish brown, similar in color to the glacial drift. They are derived both from the drift and the Precambrian bedrock, and are typical graded-stream deposits, generally gravelly at the base and sandy or silty at the surface. The gravel has a wide range in size and commonly contains large boulders that have been eroded from till as the streams cut through it. SURFACE FEATURES OF THE ONTONAGON PLAIN The surface irregularities on the Ontonagon Plain and their origin are of particular importance in relation to the drainage network because they controlled, in part, the initial drainage pattern that formed as the glacial lake withdrew from the area. The most important of these features are the glacial grooves and the glacial- lake shorelines. Other features of the plain that have had less influence on the drainage are kettle holes and beaver ponds. Glacial moraines occur in the Copper Range, but none have been found on the Ontonagon Plain itself. GLACIAL GROOVES 0R FLUTES Prominent grooves cross the Ontonagon Plain and are evident in most of the more or less level areas. The grooves have not been previously described in the Ontonagon area but they were noted by W. S. White (oral commun., 1957) who interpreted them correctly as glacial grooves. Somewhat similar grooves, or flut- ings, of about the same wavelength have been described in Alberta, Canada (Gravenor and Meneley, 1958). The grooves in the Ontonagon area are not developed equally but are more distinct and densely distributed in the inner part of the plain. They become indistinct downslope toward Lake Superior. Grooves are com- pletely lacking in low swales, as for example along the Ontonagon River and the Flintsteel and F iresteel Rivers. In these areas they are presumably covered by lake sediments. The best display of grooves is in the area to the east of White Pine on the lakeward side of the Algonquin B6 shore of Lev’erett which crosses the plain at about 940 feet. This set of'grooves consists of a series of long parallel ridges and troughs that are a regular distance apart, much like a series of waves. The average wave— length measured in two traverses between White Pine and the Cranberry, River is 380 feet. Grooves unmodi- fied by running water are generally less” than 10 feet deep, but many grooves are deeper because they have been eroded and scoured by water. Not all the grooves in the Ontonagon area trend in the same direction (fig. 3); rather, they reflect the direction of motion of the ice. The fact that in some places one set of grooves crosses another suggests that groups of grooves were formed at different times by ice that moved in- slightly different directions and that younger sets obliterated older ones. The grooves are not continuous nor all of the same length. Some are as much as 3 miles long; others are less than a quarter of a mile long. ‘ Similar grooves or glacial flutings in Alberta, Canada, have been studied by Gravenor and Meneley (1958). Their data Show that the flutings are parallel to the direction of ice motion. The Alberta grooves resemble SHORTER CONTRIBUTIONS 89°30’ TO GENERAL GEOLOGY closely the Ontonagon grooves and their wavelengths are similar. The mode of the wavelength frequency distribution in Alberta is between 350 and 400 feet; in the Ontonagon area the mode is approximately 350 feet and the average is 380 feet. Gravenor and Meneley (1958, p. 7 24) made microfabric studies of the till be- neath the Alberta flutings and found that at depths less than 5 feet the particles had a strong preferred orienta- tion parallel to the flutings. At depths of 11 feet and below the orientation was not parallel to the flutings. In the Ontonagon area, F. Brandtner in 1960 made fab- ric analyses of the till at three localities and found that the pebbles in the upper till showed a strong pre- ferred orientation roughly parallel to the grooves (fig. 3). In order to further establish the direction of ice mo- tion in relation to the grooves, the writer searched bed— rock outcrops in the Copper Range and the Porcupine Mountains for glacial striations. Nine localities were found within the area shown on plate 1 that give clear evidence of the direction of ice motion (fig. 3). Such glacially polished bedrock surfaces are not com- mon on the volcanic rocks because of the rapidity of 89°15’ 89°45’ SUPERI / [)1/ 1/! Kill, t) ”I! MK I I ’n I' . MI ”I N Z [/l I“ ill/\w I) l/// /} //l \ ”’/ fly mil/l I * (WWW/fl) PORCUF’INE MOUNTAINS // //“/ .'//// // %/ 46°45’ /’ 'llli/llli;fJ/wa ’ f;— \ \ WOntonago ,\ WE (.W 1/}, /// I l) r. i / [if I l 1 HI I ’1 In (l ’/ ,% /// /// //\l\ _ ///{///%”/m g \ l illllllh \ W)“ ‘ \ 5 MILES EXPLANATION / / V /// Glacial grooves /p’3\\\,\Shorelines of glacial lakes / Glacial striations / Orientation of stones in till FIGURE 3.—Ontonagon area showing minor surface features including glacial grooves, glacial-lake shorelines, and localities where direction of ice movement could be determined. POSTGLACIAL DRAINAGE, STREAM GEONIETRY, ONTONAGON AREA, MICHIGAN weathering, but they can be observed in some places by stripping off the cover of soil and moss, if it is thin enough, and washing the rock surface with water and a scrub brush. In every case where striae or grooves on the rock surface were found, their directions corre- sponded to the directions of the large grooves on the surface of the till plain nearby. As shown in figure 3, in some parts of the Ontonagon Plain the grooves are missing. In some places, as along the Flintsteel and Firesteel Rivers, this is because the surface is underlain by thick lacustrine deposits that are younger than the grooves. The grooves, if present, are buried. The grooves were modified slightly by shore erosion and deposition after the ice withdrew. The thin cover of lake sediments that occurs on most of the surface, however, is too thin to obscure the grooves. The con- tinuity of the ridges, however, is interrupted in various places and the grooves are weaker along the old shore- lines, if not entirely obliterated. GLACIAL LAKE SHORELINES The traces of at least seven glacial lake shorelines were found in the Ontonagon area (fig. 3). These features are best observed on aerial photographs on which changes in texture of the vegetation or breaks in con- tinuity of various surface features are apparent. The shorelines are most conspicuous in bedrock areas where gravelly beaches and nips cut in the rock are well pre- served. In till areas, the shorelines are rather obscure and some cannot be found on the ground. Locally, gravelly spits, bars, and beach ridges are used as sources of aggregate and road metal. Highest Lake Duluth shore—The best most complete of the higher shore lines found by the writer in the area shown on plate 1 corresponds to What Leverett (1928) called the highest Lake Duluth shore. In most of the area it is marked by a prominent nip with sand and gravel deposits in front of it. Spits, bars, and even tombolos are well formed. The shore is readily traced on aerial photographs and also can be seen on the ground without difficulty almost anywhere along its trace. Ac- cording to Leverett (1928, p. 62), the Lake Duluth shore is tilted and rises in elevation toward the east from 1,163 feet in the Porcupine Mountains to 1,192 feet near Greenland; it rises sharply near the Ontonagon River. The writer’s tracing of this shore, based mostly on aerial photographs, corresponds closely to Leverett’s descrip- tion and the shore apparently rises from a minimum of 1,150 feet in the Porcupine Mountains to 1,190 feet north of Greenland, a difference of 40 feet. There are at least two‘ shorelines between the Lake Duluth and the Lake Algonquin shore of Leverett. B7 Only the lower of these two can be traced across the Ontonagon Plain and even it is discontinuous. It is most distinct south of White Pine where gravel has been mined from it for use on the railroad. This shore- line rises from 1,020 feet to 1,060 feet from west to east. Lake Algonquin shore of Leverett.—A fairly distinct shorelines occurs at an altitude of 920 feet in the Porcu— pine Mountains and at 960 feet in the Greenland—Rock- land area. This shoreline corresponds to the one that Leverett (1928) referred to as the Lake Algonquin shore. It is well formed and fringed by gravelly beach deposits from the Porcupine Mountains eastward to and beyond White Pine. It is also well formed in the Greenland-Rockland area, but between these places it is indistinct and, though traceable on aerial photographs, it cannot everywhere be found on the ground in the course of a traverse. It crosses and obscures the glacial grooves, however, and this is the best means of recoge nizing it (fig. 3). At least three lower shorelines cross the plain between Leverett’s Algonquin shore and the Lake Nipissing shore. The highest of these is fairly well marked by its effect on the glacial grooves. It rises from 715 feet in the Porcupine Mountains to 755 feet east of Ontonagon. Lake Nipz'ssz‘ng 8h07‘6.—A very conspicuous shoreline occurs at an elevation between 615 and 620 feet above sea level or approximately 15 feet higher than the pres- ent shore of Lake Superior. The base of the nip, is covered by bars and dune sand, and the writer did not determine whether this shore is tilted within the area studied; the tilt, if present at all, must be slight. The Lake Nipissing shore is much better developed than any of the others, including the highest shoreline of Lake Duluth. The nip or scarp is very well marked and in places is 20 feet high, though generally only 5 to 10 feet high. Furthermore, a fairly wide belt south of the cliff is mantled by sand presumably blown from the shore below. On the lakeward side, the shoreline is fringed by a wide belt of beach ridges and dunes that form a sandy complex far larger than any of the older beaches ' AGE OF THE SHORELINES In the late Pleistocene and Recent chronology of the ancient Great Lakes outlined by Hough (1963), the Valders ice retreated rapidly and glacial Lake Duluth must have been short lived. The maximum advance of the Valders ice sheet occurred at about 11,000 B.P. (before present). The ice retreated from the area and formed the ice-marginal Lakes Ontonagon and Duluth; it eventually reached a level below the present shore of Lake Superior when the water of the basin drained over a sill at Sault Ste. Marie into lower lakes to the east. B8 The age of these lower lakes is thought to be about 9,500 B.P. (Hough, 1963, p. 103). Thus the ice must have retreated from northern Michigan, including the On- tonagon area, and the glacial Lake Duluth drained within a period of less than 1,500 years. Samples of wood collected by F. Brandtner in 1960 provide a local chronology in agreement with Hough’s analysis, and show that within the Ontonagon area itself the retreat of the ice and draining of the lake probably occupied an even shorter time. Brandtner collected samples of wood from a kettle-hole swamp near White Pine uncovered by a power shovel during excavation for a tailings pond of the White Pine Cop- per Co. The kettle is near Caribou Creek in sec. 2, T. 50 N., R. 42 W., at an altitude of about 880 feet, 280 feet below the high stage of Lake Duluth. A log , found in the upper clayey till beneath the kettle hole yielded a carbon-14: date of 10,230:280 years B.P., as determined by Meyer Rubin of the U.S. Geological Survey (W—964). Dates obtained from a log found within the kettle and from gyttja at the base of the kettle were 9,600i350 B.P. (W—965) and 9,500i350 B.P. (W—1150) respectively. The date of the till pre- sumably represents a minor readvance of the Valders ice. The kettle could not have contained a fill of gyttja until the water of Lake Duluth had withdrawn from the area. These dates indicate that Lake Duluth may have formed and drained in a period less than 780 years. The rapid withdrawal of water from the Ontonagon Plain is also suggested by the uniform tilt of the traces of the successive shorelines. All the older shore lines are tilted nearly the same amount and they rise about 40 feet between the Porcupine Mountains and the area of the Ontonagon River. By contrast, the Lake Nipis- sing shore has no measurable tilt within the area studied; furthermore, none of the older shorelines are prominent features as compared with the Lake Nipis- sing shore. The evidence on the Ontonagon Plain seems to be in close agreement with the chronology of Hough (1958, 1963) but not with that of Leverett (1928). Leverett believed that Lake Algonquin included the area of the Superior basin and formed a prominent shoreline in the Ontonagon area at 920—960 feet. Although a shore- line does cross the plain at that level, it is tilted the same amount as the other shorelines, including the high shore of Lake Duluth. The kettle hole from which the date of 9,500 B.P. was obtained lies below Leverett’s Algon— quin shore. Thus, the evidence indicates that all the shorelines north of the Copper Range older than the Lake Nipissing shore should be assigned to the Duluth stage. SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY No evidence was found in the area by which to date the Nipissing stage, but it is well documented in other areas. The high stage presumably occurred at 4,200 B.P. (Hough, 1963). KETTLE HOLES Small subcircular kettle holes, which have been re- ferred to locally as “bathtubs”, are abundant on the Ontonagon Plain. They are especially numerous in places where the glacial grooves are best developed. Very few of these kettles are more than 100 feet in diameter and most are about 50 feet or less. Some are ponds or lakes, others are bogs, and still others are dry sediment-filled depressions. The larger kettles are readily visible on aerial photographs. Their position at the top of the till sequence and the fact that they contain undisturbed lake sediments indi— cate that the kettles must have formed during the last advance of the ice, probably contemporaneously with the grooves; they must also have formed either under the ice or under water. They probably are holes left by ice that became entrapped in the clayey till and later melted. These features are not large enough to have influenced the development of the drainage. BEAVER PONDS The dense distribution of beaver dams and ponds on the Ontonagon Plain suggests that beavers may have influenced the drainage development. Residents of the region classify the beavers as bank beavers and pond beavers, depending on the environment in which they make their homes. The bank beavers inhabit the in- trenched stream valleys, including streams as large as the Cranberry River. They build small dams across the channel, but these dams are low and generally do not reach from one floodplain to the other. Because beaver lodges are generally not in evidence along the large streams, the animals presumably inhabit burrows in the stream banks. Dams on the large streams are temporary and probably have no permanent influence on the channel equilibrium. The habit of some beavers of digging bank burrows along streams that are regu- larly subject to floods is apparently well known to bi— ologists and has been described by Dugmore (1915, p. 28). The pond beavers live in more permanent bodies of water on the upland area of the plain. Glacial grooves make excellent enclosing walls for lakes and beaver ponds, averaging one or two per square mile, are espe— cially abundant in the grooved areas. Often the dams are very elaborate and long—lasting structures on which shrubs and trees grow and which may enclose bodies of water many acres in size. The pond beavers live in POSTGLACIAL DRAINAGE, STREAM GEOMETRY, ONTONAGON AREA, MICHIGAN typical beaver lodges that are much in evidence at the upland sites. The pond beavers may have some influence on the geomorphic development of the landscape. It is pos. sible that some of their dams are large enough to divert small upland streams from one groove to another. Study of the aerial photographs indicates that such diversions have been rare, however, because most dams do not rise as high as the top of the ridge between the grooves. CHRONOLOGY OF EMERGENCE OF THE ONTONAGON PLAIN The chronology of events during the uncovering of the area and the changes in position of the lake shore have an important bearing on the development of the present subaerial drainage system, for the lake shore is the base level of erosion. The events that took place following the high stage of Lake Duluth are summarized below as they are now known from evidence found in the Ontonagon area, as well as from Hough’s (1958, 1963) synthesis of the history of the Great Lakes: 11,000—10,000 B.P.~Valders ice retreated from the area and formed the ice-marginal Lake Duluth that covered the Ontonagon Plain to an altitude of 1,100— 1,200 feet above present sea level. This open water body had sufficient fetch for wave energy to cause considerable shore erosion in a short period of time. 10,000—9500 B.P.—The level of the ice-marginal lake was lowered and remained for short periods at various levels. At the end of this period the shore was probably somewhat below that of the present Lake Superior. 9500—4200 B.P.—The events in the Lake Superior basin during this interval are not well known, but presumably in the Ontonagon area the surface of the plain was tilted, rising to the northeast, as a consequence of the unloading of the ice. At the same time the level of the lake rose both because of the tilt and because of changes in the outlets of the lower lakes of the Great Lakes basin. At 4,200 B.P. the shoreline of Lake Nipis- sing in the Ontonagon area must have been at about 615 feet, 13 feet higher than the present level of Lake Superior and a strong nip, or escarpment, was cut back into the Ontonagon Plain. 4,200 B.P. to present—The level of the lake was lowered slightly; a broad beach and series of bars were built lakeward from the Nipissing scarp. The lake shore reached its present level of 602 feet. Some tilting occurred after the Nipissing stage, but the effects were not large enough to be observed in the Ontonagon area. EVOLUTION OF THE STREAM VALLEYS The evidence indicates that the area must have been free of lake and glacial water and exposed to erosion B9 by streams as long as 9,000 years ago. Except for the narrow belt below the Lake Nipissing shore, it has been subject to subaerial erosion ever since. The gyttja and buried log in the kettle hole near White Pine (p. B8) show that the plain was very soon covered by dense vegetation when the water receded. Probably the only areas free of vegetation were the larger runoff channels where floods prevented cover growth. The lack of evidence of piracy (pl. 1) indicates that little change in the drainage pattern has taken place since the drainage system formed. Most streams follow the pattern of grooves, and captures have occurred only where the larger valleys have grown laterally and en- gulfed smaller ones. As an example, the Ontonagon River has widened its valley somewhat, apparently as its meander belt shifted, and has captured Mill Creek. Smaller valleys, however, especially where the grooves are most dense, show little evidence of this process. Tributaries of the Floodwood River, for example, parallel the Potato River for miles without joining, even though the Potato is a much larger stream. Stream junctions more commonly occur where initial irregularities on the surface of the plain controlled the direction of flow. Deer Creek (the upper part of the Potato River) serves as a striking example. This stream flows down a series of grooves that are oblique to the general trend. On its west bank it picks up a number of smaller tributaries that flow parallel to the more general trend of the grooves. The initial drainage pat- tern must have been determined in detail by the surficial irregularities of the emerged sublacustrine plain. Valley cutting has not generally proceeded headward from the lake; instead, the drainage channels have deepened all along the course an amount roughly pro- portional to a function of drainage area, and hence dis— charge. Rudimentary channels generally begin 1 to 2 miles below the head of a groove where drainage areas are only 1/2 to 1 square mile in size. Because the grooves are shallow and on the average are 380 feet wide, the smallest stream channels form only shallow and narrow gullies that occupy a small part of the floor of the groove. Gradually a valley begins to form which con- tains a meandering channel. Valleys more than 5 miles long may have widths equal to the wavelength of the original groove in which the valley was formed. In the uppermost reaches the small channels are grassy and floored with fine-grained material. The size of the bed material gradually increases as the channel de- velops. Argentine Creek, for example (see. 18, T. 50 N., R. 41 W.), 2.4 miles from the source, is intrenched in a steep-walled ravine 60 feet wide but only 6 feet deep. The stream channel within the ravine is 6 feet wide and less than a foot deep. The bed material in the channel B10 is medium to fine sand. At 3.7 miles from the source the same stream is in a valley 120 feet wide intrenched 8 feet beneath the groove floor. The stream channel is 12 feet wide, 2 to 3 feet deep, and has coarse sand on its bed. The increase in valley depth as the drainage area be- comes larger is shown graphically in figure 4. The inference is made from this relationship that, in general, the stream valleys have not developed by head- ward erosion from base level, that is, from Lake Su- perior, but have gradually deepened through time in proportion to a function of the runoff carried by the channel. 7 In the valleys of the streams that enter Lake Su- perior, no evidence is found of nickpoints that migrate upvalley. . Instead, as shown in figure 5, the profiles of the smaller streams show scarcely any intrenchment at all, even within a few tenths of a mile of the low escarpment south of» the lake. Larger streams have slightly convex profiles as they approach the lake. The largest of the streams, such as the Little Cranberry River which has a drainage area of about 6.5 square miles, have concave—upward profiles. The data sug- gest thatonly the smallest streams have profiles that might be considered immature. Theoretically, the pro- SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY files of even the smallest should be concave upward if ~ the form is related to increasing discharge and if com- plete equilibrium of form has been achieved. The smaller streams, however, have not kept pace with the lowering of the lake level and the cutting back from the shore that occurred during the Nipissing stage. The Floodwood River profile, for example, has a slight con- vexity as it approaches the lake. When the relation of the profile of a valley to its base level is considered, the evidence can be interpreted to mean that the effect of base level on the forms of small valleys like these is mainly to determine the energy gradient. Even small valleys less than 2 miles long far back on the plain are about as deep as valleys of similar size close by the lake. The dominant factor in the val- ley cutting appears to be the discharge available, and valleys with drainage areas less than a certain limit (about 6 square miles) still preserve marked convexi- ties where they enter Lake Superior. DRAINAGE COMPOSITION The drainage network of the area studied is conse- quent, and few adjustments have taken place by piracy. The present drainage system must be similar to the initial drainage that developed as the lake withdrew 400_ I‘ll Illll l l l Illll ll lllll I IIIIIII _ _ ._ ._ . _ O 100'_ - ' _ : 1": ° ' 3 ,|_. r . _ a: — - — LL .2 _ O. _ ;— - . - — )— I D. 0 L5 — ..o _ >_ . LLJ :1 <10 ‘ > - _ . _ . _ _ , _ 110' JALIILII . Illlll lllllll VI lllllll 0.1 1.0 10 1007 ‘1000 2000 DRAINAGE AREA, IN SQUARE MILES FIGURE 4-.—‘Relation of valley-depth to drainage area of valleys eroded in the Ontonagkm Plain. POSTGLACIAL DRAINAGE, STREAM GEOMETRY, ONTONAGON AREA, MICHIGAN Area: 0.17 sq mi 660 — 640 - 620 — 600 Stream 0.2 mi west of Cranberry River 700 — 680 - 660 v 640 m 620 — 600 \\ \\ \ \ \ \ Stream tributary to Cranberry River on east bank 0.8 mi upstream from Lake Superior Ate Lake 680 _ 660 — 640 P 620 *- 600 Stream 0.5 mi east of Cranberry River ALTITUDE, IN FEET 680 » ‘\\\ 660 — ‘x 640 ~ 620 — 600 Floodwood River 700 — ‘\ 680 _ \\\\ Area: 6.5 sq mi 660 - 640 — \ 620 — 600 \ Lake .10 MILES Little Cranberry River 2,5 2,0 1,5 l.0 0.5 I | FIGURE 5.—Lower parts of profiles of several small streams show— ing the change in character related to the (inference in size of drainage area. Solid line indicates smoothed profile of stream. Dashed line indicates upland surface. Dots on profiles are altitudes as measured by Kelsh plotter. from the plain. The geometry of drainage patterns can be studied in various ways. The system of drainage analysis developed by R. E. Horton (1945) shows that drainage networks are organized in a regular manner and that they can be quantitatively described. In Horton’s system, streams are classified according to or- der increasing downstream: A stream 'with no'tribu— taries is a first—order stream; a stream having one or more tributaries of the first order is a second-order stream; a stream having one or more tributaries of the second order is a third—order stream, and so on. The number of streams of each order in a drainage network decreases as order number increases in .a geometric series. The average length of streams of different or- ders increases as the order increases and other char- B11 a‘cteristics, such as average slope, also change in regular geometric series. Horton showed that various ratios, obtained by simple measurements of the drainage net- work, could be used to define and describe the geometry of individual drainage networks. Strahler (1957) made use of Horton’s system of analysis, though he modified it slightly, and described some of the important variations in the drainage patterns of different geologic environments. Leopold and Miller (1956) described the relation between Horton’s drainage composition and the hydraulic characteristics of streams. More recently Leopold and Langbein (1962) have suggested that the arrangement of stream attributes 111 regular geometric series is the most probable arrangement in a system in dynamic equilibrium. The writer, interested in the shape of drainage basins, has studied the relation of Horton’s geometry to stream length and drainage area in previous work (Hack, 1957). Drainage basins increase in length relative to their Width as they become larger, and the rate of 1n- crease is much the same for drainage basins throughout large areas. In general, the length of a drainage basin measured along the principal stream increases in pro- portion to 1.4 times the six- -benths power of the drainage area. Miller (1958, p. 10) found similar values for streams in New Mexico. The relation to Horton’s geometric system indicates that certain of the attributes of drainage patterns he described, such as length and bifurcation ratios, are very stable. In the Ontonagon Plain area, because of well-de~ veloped glacial grooves, the drainage basins have elon- gate cigarlike shapes. The cigarlike basins, as shown by plate 1 are confined to the areas of the grooves and the drainage lines are arranged in a strongly parallel pattern. In areas without grooves, as in the Ontona— gon River valley and the basins in the Copper Range, the drainage is typically dendritic in plan. Some pat- terns, like that of Mill Creek, are partly parallel and partly dendritic. The analysis that follows concerns principally areas of parallel drainage. Several examples of drainage systems on the Ontona- gon Plain are shown in figure 6, obtained both from topographic maps and aerial photographs. The first— order stream channels generally can be seen only on the aerial photographs. Note also that the Cranberry River basin has the characteristic elongate shape only in the lower part where it traverses the grooved area. The relation of‘ stream number and stream length to order are shown in figure 7. The values of the various parameters defined by Horton are shown in table 1. These parameters are defined briefly below; a complete B12 SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY Lake Lakes/Lore / \ \ \ Little Cranberry River (From aerial photographs) Streams between Mineral River and Cranberry River (From topographic maps) Cranberry River (From topographic maps) WW N / o 1 2 3 4 MILES _ _ |___|__L__L___|____| lel Creek Wetgel Creek (From aerial photographs) (From aemal photographs) FIGURE 6.—Examples of drainage systems on the Ontonagon Plain. POSTGLACIAL DRAINAGE, STREAM GEOMETRY, ONTONAGON AREA, MICHIGAN 50 ._. 0 AVERAGE LENGTH OF STREAMS IN MILES m U! 0 NUMBER OF STREAMS 0! S 0.5 STREAM ORDER (INDICATED BY NUMBER) FIGURE 7.—Relation of stream order to stream length and num- ber in drainage basins of the Ontonagon Plain. A, Streams between Mineral and Cranberry Rivers; data from maps; stream orders not known. B, Little Cranberry River. 0, Weigel Creek. D, Mill Creek. explanation of these parameters is available in Horton (1945). A, area; area of drainage basin, generally in square miles. L, length, in miles, of longest stream or principal stream in drainage basin. N, number of streams. S, stream frequency; refers to number of streams per unit area, generally per square mile. Dd, drainage density; length of stream per unit area, generally in miles per square mile. B13 1., average length in miles of streams of order s. as, average drainage area in square miles of streams of order 3. 17,, bifurcation ratio; ratio of the number of streams of one order to the number of streams of the next higher order. in, length ratio; ratio of average length of streams of one order to streams of next lower order. p, ratio of length ratio to bifurcation ratio. 8, order of main stream in a given drainage basin. The drainage basins on the Ontonagon Plain are obviously well ordered (fig. 7 ; table 1); that is, the streams of different orders vary in number and length in the same manner as the streams described by Horton. Comparison of the bifurcation ratios and length ratios (table 1) with those given by Horton (1945, table 1) also indicates that they are similar. For example, the average of 10 of Horton’s length ratios is 2.3, not far from 2.5, the average of the ratios of four segments of the Ontonagan Plain shown in table 1. The bifurca- tion ratios of Horton average 2.9, whereas the ratios of the four segments of the Ontonagon Plain average 3.0. Not only are these ratios similar, but the fit of the data when plotted graphically (fig. 7) appears to be about as good as the fit obtained by Horton (1945, figs. 2, 4). The factor that distinguishes the streams on the Ontonagon Plain is a high drainage density and elongation of every order of stream. Weigel Creek on the Ontonagon Plain, for example, has first order streams that average 1.2 miles in length but only 0.21 square mile in area. Calculation from Horton’s data (1945, table 1) shows that in Esopus Creek in the Cats- kill Mountains the average drainage area of a stream 0.99 mile long is approximately 1.1 square miles. Thus, on the Ontonagon Plain, a first-order drainage basin has only one-fifth the area of a basin of equivalent length in the Catskill Mountains. The shape of drainage basins can be studied more directly by comparing the length of the principal stream TABLE 1.—0haracteristics of the drainage net of certain stream basins and groups of basins in the Ontonagon area Average Average _ Hi best Total Number Number of Stream Drainage length of drainage Bifur- Len th Basm(s) or er of area. of first-order frequency density first-order area of cation rat 0 p stream (sq mi) streams streams streams first-order ratio stream All streams between Ontonagon and Mineral Rivers (data from topo aphic maps) ...... 3 126 93 70 0.74 2. 42 2. 05 0.85 2.85 2. l 0. 74 Same, converted to h gher order 1 ________ 4 126 293 200 2. 3 3. 3 .97 .30 2. 85 2. 1 .74 All streams between Mineral and Cranberry Egggédétavigrotrgitgpgyaphic amps) ...... i 3 1:3 36 .17 2. 86 2. 45 . 86 2. 85 1.9 .61 , n 1 er or er ........ 1 . . . . . . . Littlle Eranberrty )River baiin (data from aer- 39 2 4 2 1 27 30 2 85 1 9 61 is p oogra s ........................... 4 6.5 29 3.0 . . . . 1.5 Weigel Creek lbasin (data from aerial photo- 20 7 52 07 2 86 3 0 0 graphs)___-..._ .................. .. .......... 4 5.7 28 19 4.9 5. 6 1.2 .21 3.28 2. 58 .79 Mill Creek basm (data from aerial photo- graphs) ____________________________________ 5 9. 2 119 88 13 6. 6 . 31 . 05 3. 15 2. 5 . 80 1 I nasmuch as the smallest streams are not indicated on the topographic maps, it is assumed that the smallest streams shown are on the average of the second order rather than the first, and Horton’s parameters are recalculate . 748—958 0-65—3 B14 with the drainage area of the basin. It has been shown that in erosionally graded landscapes in the eastern United States such as the Shenandoah Valley of Vir- ginia, the length of the principal stream, within fairly narrow limits, is equal to 1.4 times the six-tenths power of the drainage area. In other words, the average stream with a drainage basin 1 square mile in area is 1.4 miles in length, and increases in length at a rate that is slightly greater than the square root of the drainage area and approximates the six-tenths power. This re- lationship is quantitatively related to the parameters in Horton’s system (Hack, 1957, p. 63—67). By using Horton’s equation relating drainage density to other properties of the drainage network, the length of the principal stream of a drainage basin of order s can be expressed by the equation : L8=erls_1 (1) The area of a drainage basin of order s can be ex- pressed by: Lfl:i p—1 (m Thus the relation of length to drainage basin area is a complex function of the drainage density, bifurcation ratio, and length ratio (Hack, 1957, p. 66—67). Trial of different values for the variables shows that the ex- ponent in a general equation expressing basin shape, L=bA" (3) is determined by the quantity p, the ratio of the length ratio to the bifurcation ratio. The value of the coeffi- cient, b, is related indirectly to the drainage density, but also to p, for it is equal to the quantity L/A” or ll/ai'. Figure 8 is a graph on which the length-area relation- ships for several hypothetical drainage basins have been plotted. The values of the exponents and coeflicients are determined by assuming certain values of a1, ll, 7'1, and r1, as shown in the inserted table. We can thus observe the overall change in the basin shape as the basin grows in size, as reflected in the slope and position of the curves on the graph. Curve 2 of figure 8, with a slope of 0.5, serves as a good reference line, for when the exponent n has a value of 0.5 the drainage basin does not change in overall shape as the basin becomes larger, that is, the length of the basin as measured along the principal stream is always proportional to the square root of the drainage area. If n is larger or smaller than 0.5 then the shape of the drainage basin along a given principal stream must change in a downstream direction. If n is greater than 0.5 then larger basins are narrower rela- tive to their lengths than smaller upstream basins, As=alrlf SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY whereas if n is smaller than 0.5 larger basins are wider relative to their lengths than smaller ones. The inter- cept at a drainage area of 1 square mile is equivalent to the coeflicient and is determined by the absolute values of length and area of the average basin of a given stream order. The construction demonstrates that differences in the shape of drainage basins theoretically can be con- trolled by difl'erences in bifurcation and length ratios having a wide range in value. If the bifurcation ratio characteristic of a drainage network were very large, basins would tend to become more circular in shape with increasing size. If the length ratio, on the other hand, were large, drainage basins would tend to become narrower with increasing size and large basins in such a network would be very narrow and cigar-shaped. The evidence, however, indicates that the ratios 7'; and 7",, are very conservative and tend to have values like those of curve 4 (fig. 8) in which p is 0.75, and which represents a drainage basin in the Shenandoah Valley (Hack, 1957, p. 66). Analysis of the streams on the Ontonagon Plain by Horton’s method has shown that the ratio p is similar, in that it averages 2.5: 3.0 or 0.83. Although the extremely elongate basins of this area do become slightly more elongate as they enlarge, the rate of change is very little greater, if any, than that in most other, more normal drainage networks, and first-order drainage basins are nearly as elongate as third or fourth-order basins. The actual dimensions of drainage basins on typical grooved areas on the Ontonagon Plain are shown graph- ically in figure 9. The exponent of a curve drawn by inspection through the points is 0.6 and the coefficient is 3.6. Obviously both small and large basins tend to be very elongate and the elongation becomes slightly greater as the basin enlarges. Streams like the Gran- berry and Potato Rivers, that originate south of the grooved area but cross it, follow quite a difl’erent curve, as shown in the graphs. The upper part of the Gran- berry River basin with a drainage area of about 10 square miles is like any ordinary drainage basin and has a dendritic drainage pattern. In this part of its course the principal stream follows the curve L=L4AM. m) Where the drainage area becomes 10 square miles, however, and the river is a fifth-order stream, it enters the grooved area and follows a narrow corridor through the grooves to the lake, picking up no tributaries any larger than second order. POSTGLACIAL DRAINAGE, STREAM GEOMETRY, ONTONAGON AREA, MICHIGAN B15 100 _ U) _ Lu 2' E Z 2- 10__ < _ Lu _ a: ’_ _ U? _ _1 < _ E o __ E 0: fl- ._ n. O a“ I. 1.0 _ / ,_ _ o _ Z _. LIJ -| _ 0'1 I lllllll IIIIIIII IIIIIIII lllllll 0.1 1.0 10 100 1000 AREA, A, OF PRINCIPAL STREAM, IN SQUARE MILES Curve ll (11 n 75 p b n l 0.5 0.33 5.0 2. 5 2.0 1.25 0.85 2 .5 .33 2.5 5.0 0.5 .86 .5 3 .5 .33 2.0 10.0 .2 .78 .29 4 .35 .077 2.4 3.2 .75 1.50 .62 FIGURE 8.-——Possible relations of stream length to drainage area in hypothetical drainage basins having difierent assumed values of length, area, length ratio, and bifurcation ratio. calculation. Leopold and. Langbein (1962, p. A14—A17) have shown that the geometry of natural drainage can be explained simply as ordered in a random manner. They show that if streams of each difi’erent order are spaced an even distance apart the streams will, on the average, join at distances that are proportional to the stream order. Leopold and Langbein further show that the length ratio must range between 2 and 4, and that natural streams do have length ratios between these limits. Furthermore, by using a random-walk tech— nique they construct a model drainage system on the basis of this concept. The stream length of the result- ing system varies proportionally to a power function of the drainage area close to six—tenths. Dots refer to the values of L and A obtained by The drainage network on the Ontonagon Plain, in spite of its unusual plan, behaves in' accordance with the requirements of the theory. The grooved area of the plain may be thought of as a homogeneous environ- ment in which a steep slope toward Lake Superior and a strongly grooved surface texture have together im- posed a constraint on the formation of the network and have tended to produce long narrow drainage basins. In other words, when the plain emerged from the lake, a stream of water flowing down the slope was required to travel an extra long distance before it by chance merged with an adjacent stream. Nevertheless, the bifurcations were arranged in a random manner as in any other drainage network. B16 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY / /// _ /9’ V// J //‘ / // l a) 107 /,’ *PF 5 — o7 // l. _ _ /1 2 E / o o I. Z _ / - _ 4 0’ ’ I (DO / .// l— / , (D — 31 / O / z o/ 6 / Lu 1 0’ o / .1 0 / / T O4// //0 / x o3 97/ // _ //01 x. 2 /o/ // 4 O// 1 // 3/ l/ 1 1/63 /" 1033/ I llllll llllll [III]! "0.1 1.0 10 100 DRAINAGE AREA, IN SQUARE MILES FIGURE 9.—Re1ation of channel length to drainage area of streams on the Ontonagon Plain. 1, Weigel Creek; 2, Bear Creek; 3, Floodwood River; 4, Miles Creek; 5, Stony Creek; 6, Duck River; 7, Argentine Creek; 0, localities on streams originating in grooved areas; 0, localities on Cranberry River. Actually, the drainage system on the plain is homo- geneous and randomly arranged only when examined as a whole. The minute details of the network were presumably determined by specific geologic causes and events. The drainage network began to form as Lake Duluth drained and the lake water receded from the grooved surface of the till and lake sediment. Be- cause the gradient of the plain was steep, drainage could only flow down the slope and must have been confined in the grooves. Retreat of the lake was neither regular nor immediate and the shoreline stood for short or long periods at various levels on the plain. Each stand of the lake provided an opportunity for erosion at the lakeshore of part of the ridge between grooves, and thus could have formed a gap through which a stream of water running down the plain could escape into the next groove. Such gaps are in addition to gaps present where a groove ends. The drainage pattern probably was formed by the joining of streams through such gaps as the lake withdrew. Some additional junc- tions may have been caused by beaver dams, which are still numerous on the Ontonagon Plain. If we knew the exact configuration of the plain as it emerged from the lake and the location of all the beaver dams in the past history of the area, we doubt- less could explain the location of every bifurcation. We cannot, of course, and because the bifurcations are spaced across the plain in a complex manner without apparent order, we say they are distributed in a random manner, which is equated by Leopold and Langbein (1962) with maximum probability. By using their random-walk technique, we can set up limits or con- straints and construct a model drainage pattern that resembles the actual pattern of the Ontonagon Plain. First the assumption is made that the plain is crossed by grooves 0.1 mile apart and that water must flow down every groove. In the first mile there is no chance for the water to escape, but in every succeeding mile downslope the stream has an even chance to escape to the right, to the left, or to continue down the groove. Another way of stating the conditions would be to say that the drainage density is 10 and every first-order stream has a length of 1 mile. It is not necessary to state the manner in which the stream bifurcates, for this is left to chance, the one limitation being that a stream must be 1 mile long at the first bifurcation and that it does not have another chance to bifurcate for another mile. A sample of the resulting drainage sys- tem is shown in figure 10. When the length of the principal stream is compared with the area of the drain— age basin at various points in the model, the two quan- tities are found to be related roughly according to the equation L=4A°~67, not unlike the actual equation for the Ontonagon Plain. A similar drainage network could have been constructed by using somewhat differ- ent limitations or constraints. For example, we might have assumed that at every 0.1 mile instead of at every mile, each stream has a chance to flow to the right or POSTGLACIAL DRAINAGE, STREAM GE-OMETRY, ONTONAGON AREA, MICHIGAN l I I I I I I I I I I I I I I I I I I I I L ._-__-._ r__-- 1 f----————C.._._- - l ._-__.__..._._ £-______ _--__-.... -____.__._-______ l--.__ __ -__-___‘_____-__ __._.___ —_._______.____ .___-___.r____.__ FIGURE 10.—Mode1 simulating the drainage network of the On- tonagon Plain, constructed by the random-walk technique. B17 the left or continue downslope. The chances, however, would be uneven, so that the probability of the stream turning to the right would be one-twelfth, to the left one-twelfth, but to continue downslope ten-twelfths. The conditions that determined the natural drainage network were similar to those in the model. Geologic conditions determined that a stream must flow a long distance downslope before it could join another, and the wavelength of the grooves determined that the spac- ing of the streams must be relatively close. Because the locations of the bifurcations are randomly arranged, as in any other drainage network, the resulting drain- age pattern is attenuated in a direction parallel to the grooves, as though stretched out in one direction. Pre- sumably, any geologic or topographic factor that re- . stricts or impedes the development of channels in a systematic way would operate in a similar manner. Thus on a very long slope that is unusually steep, we might also expect an attenuated drainage network. Similarly, a trellised drainage pattern is in efl'ect an attenuated drainage network caused by the presence of resistant beds of rock. If the resistant beds are far apart, the attenuation may affect only the streams of higher orders, and not the smaller streams. In the general equation relating stream length to area, L=bA‘", where L is stream length, A is area, and b and n are constants, the exponent It must always have approxi- mately the same value in any homogeneous area for its value 1s determined by the laws of probability and not by geologic factors. The coefficient 1), however, is a measure of the attenuation of the drainage network and may be affected by any geologic or topographic charac- teristic that tends to direct the flow of water uniformly. LONGITUDINAL PROFILES Studies of the equilibrium conditions in stream chan- nels have shown that many variables are involved in determining the form of longitudinal stream profiles. Classic geomorphic theory in America assumes that the concave upward form is a result of a graded condition that evolves as the landscape is eroded toward base level (Hack, 1960). Increasing discharge has generally been assumed to be an important factor. Leopold and Mad- dock (1953) in their analysis of the hydraulic geometry of streams showed that at least seven variables are in- volved in a solution of the problem. They are down- stream changes in channel slope, depth, width, load, velocity, and roughness, and discharge. Some of them are closely related to the geology of the area through which the stream passes. B18 The actual forms of the profiles in small streams com- parable in size with those in the Ontonagon area have been studied in the humid eastern United States (Hack, 1957). It was found, as might be expected, that the profile itself is closely dependent on the geology, prob- ably through the interaction of some of the other vari— ables. Most of the profiles studied could be expressed by integrals of the function SocL” where S is channel slope and L is channel length or distance from the source of the stream. Because most of the profiles are concave, the constant 3/ is generally negative and ranges in value from higher than —0.5 to lower than — 2.0. In some streams, in which the size of the bed material is uniform in a downstream direction, the value of y is —1.0; the profile is then a simple logarithmic curve H = 0 -— ylogeL, where H is altitude, L is channel length, and 0 and 3/ are constants. One stream studied, the North Fork of the Shenandoah River, follows such a curve for more than 100 miles (Hack and Young, 1959, fig. 3). Samples of the bed material of this stream, as of others having this type of profile were approximately the same in grain size all along the course. The beds of these streams, although smooth, are being eroded and contain many rock outcrops in the channel. The loose bed material is of the same size because it is locally derived or has been transported only a short distance and is of similar resistance. In such streams the competence must be more or less the same all along the course. It has been shown (p. B9) that the profiles of many of the smaller streams that enter Lake Superior are affected by the escarpment at the lake edge and have convex upward profiles. Only the larger streams have absorbed the effect of the escarpment and have concave profiles. It is expectable, then, that the profiles of many streams, especially the smaller ones, reflect the initial configuration of the sublacustrine plain and have little curvature. The channel slopes of some larger streams, however, have been reduced to much below the initial slope. The Ontonagon River has been reduced to a grade of 3.2 feet per mile where it crosses the grooved plain. The grades of the Cranberry and Potato Rivers in parts of their courses are as low as 12 feet to the mile. These grades are much less than the initial slope of the lake plain, which averages 50 feet to the mile, and they suggest at least some profile evolution. From the values for channel slope at the sample locali- ties shown in table 2, we can construct a model stream profile that represents the average for all the streams in the area. The slopes are plotted against the stream SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY length in figure 11. This model profile is not intended as a representation of all or any of the streams. Its purpose is to clarify the relation of length to area as it affects slope. The field of scatter is irregular and no definite rela- tion between the two quantities is apparent. An un- biased line through the points was drawn both by the method of least squares and by multivariate analysis (Snyder, 1962). Both methods gave the same line having the equation S=146 L'°-73, where S is channel slope in feet per mile and L is length in miles. The integral of this variable is a line that represents a model average-stream profile for the grooved plain, and has the equation H= 0 — 540L-27, where H is the fall in feet from the source of the stream and 0 is a constant. Such a profile has only a slight curvature. At 5 miles the gradient would be 45 feet per mile and at 10 miles it would be 27 feet per mile. This degree of curvature is not very high. The profiles of individual streams, of course, are all different from this average. The smaller streams, like Bear Creek, have profiles without appreciable curvature (fig. 12). The profiles of the larger streams are also irregular and it cannot be shown that there is any systematic downstream rate of decrease in slope representative of all of them. Streams such as the Cranberry and Po- tato Rivers, however, are large enough to have a marked concavity (fig.12)., The irregularity of the profiles is probably due partly to the effect of the bedrock through which the stream flows and partly to irregularities in the rate at which drainage area and discharge increase downstream. The Cranberry River, for example, has a reach about three miles long in which the lower valley walls are bedrock and rugged enough in appearance to be called a gorge. The profile is so much affected by the bedrock that the gradient increases to 50 feet per mile, whereas it is less than 30 feet per mile in the till area upstream and only 16 feet per mile downstream where the valley is also in till. The marked affect of the geology on the slope of other streams is reflected in the relation of channel slope to drainage area (fig. 13). The channel slopes at most localities in bedrock reaches are much higher than in till reaches. The fact that the gradients have become adjusted to the presence of bedrock in the valley walls is, of course, an indication that there has been consider- able gradation of the river profiles. The effect of the relation of drainage area and length on the channel slope is also shown by figure 13. Note that the average slope in a drain-age area of 1 square mile is approximately 50 feet per mile. A similar co- efficient was obtained for streams in the Shenandoah Valley of Virginia, in limestone and shale, and in fact POSTGLACIAL DRAINAGE, STREAM GEOMETRY, ONTONAGON AREA, MICHIGAN \ 100 L3 . E _ a: . g - .3- . E . LLLJ _ C . O z '.' E ' ' . 010 a‘ : - ' _' __ LLJ z _ <2: _ 5 — - \ _ . 1'0 illllll I l|l|| 1 lllllll 0.1 1.0 10 100 STREAM LENGTH, IN MILES FIGURE 11.—Relation of channel slope to channel length at localities in the Ontonagon area (data from table 2). Equation of line: S:146L'°"” TABLE 2—Measurements in stream valleys at localities in the Ontongon area B19 Estimate of bed-material size (mm) Area Length Slope Width Depth Valley Valley No. Stream (sq mi) (mi) (ft per (ft) (it) Frag- Geo- Phi depth width mi) ments in metric Phi stande (it) (ft) sample mean mean devia- tion Stony Creek ...................................... 1. 71 5. 0 32 ________________________________ Mineral River. _ . 11.45 13.0 12. 5 36 3. 5 12 ................ Tolfre Creek ___________ _ 1. 82 4. 4 36 14 2. 5 25 18 85 West ranch Duck Rive . 2. 36 4. 2 33 12 3. 5 21 20 90 East Branch Duck River. _. .. 1. 2 3. 5 67 12 2 0 l . 5 25 110 Argentine Creek ______________ . 47 2. 4 67 6 0. 8 l . 25 6 60 Halfway Riven... ...... 1.17 1.6 62 6 2 0 1 1.0 28 100 Argentine Creek. _ ........................ .87 3. 7 33 12 2. 7 l 1.0 8 120 Ha fway River .................................... 2. 00 3. 5 57 15 4. 5 47 40. 0 —-5. 3 2. 5 19 180 Iron River ________________________________________ 97. 4 20. 3 24 143 14. 0 56 52 — 5. 7 1. 0 100 900 Little Iron River .................................. 13. 3 6. 6 133 98 5. 2 51 80 —6. 3 2. 2 80 275 Cranberry River .................................. 16. 7 11. 3 50 67 6 5 52 52 —5. 7 1. 5 80 195 Argentine Creek .................................. 3. 16 7. 8 44 18 3. 0 42 36 —5. 2 1. 9 22 160 Little Iron River .................................. 13. 6 7.0 40 46 5.0 40 52 —5. 7 2 0 80 160 West Branch Iron River .......................... 33. 6 16.9 16 100 6.0 94 26 —4. 7 1. 6 80 500 _____ do..._..___............-.-.-._._._....__..-.-.- 50 16.0 26 115 8.0 45 110 —6.8 1.8 100 550 . Mineral River .................................... 6 6 7. 0 100 35 5. 0 22 84 —& 4 2 5 30 Cranberry River .................................. 17. 3 15. 2 33 45 4. 0 50 64 —6. 0 1. 1 80 1, 000 _____ do......_....._.__.......__.__......-.-__...... 17.8 15.8 31 55 6.0 50 26 —4.7 .9 _-...... ........ 17.4 15. 4 50 —4. 9 1. 5 ________________ 18. 2 16.9 30 —7. 0 1. 9 75 l, 200 18.1 16.6 50 —4. 7 .8 80 1,100 ..... do...._......___....._..._....___...-.-.____-._ 19.1 18.5 33 —5.5 .7 60 1,400 Potato River ...................................... 21. 5 14. 6 —6 0 1. 4 80 700 118 .......... do ......................................... ‘.. 21.7 15. 4 —3. 5 1. 5 75 1,400 127 _____ Deer Creek ....................................... 15.3 8.8 . .......... 40 132 ..... Cranberry River .......................... 14. 0 7. 6 6 70 144 .......... do ............. 16. 1 11. 5 40 166 _____ Firesteel River ................ 83. 8 21. 0 201 _____ West Branch Ontonagon River 183 30 202 __________ do. 3 ___________________________ 590 52 203 _____ Middle Bran ch Ontonagon River 3 305 45 204 _____ East Branch Ontonagon River 3. _ _ . . 183 25 205 ..... Ontonagon River ................................. 1, 496 58 1 Average size estimated by comparison with standard reference samples. 3 Rapid estimate by measurement 016-10 fragments chosen at random. ' Locality not visited. B20 1000 I SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY I I I 900 - 800 — 700 _ Bear Creek 600 I I I I I , I I 800 | I 700 — Lower Potato River 600 1300 ALTITUDE, IN FEET 1200 — 1100— 1000 — 900 — 800 — 700 _ Cranberry River 600 I ' Upper limit of main grooved area _ VaIIey in bedrock r—h—‘flfi O 2 4 6 8 10 12 14 16 18 20 DISTANCE FROM SOURCE, IN MILES FIGURE 12.—Longitudinal profiles of part: of three streams in the Ontonagon area. the slope-area diagram for such streams in that region is very similar (Hack, 1957, fig. 16) but the slope length relationship is quite different in the two regions. In the Shenandoah Valley the slope at a stream length of 1 mile is 80 feet per mile, about 1.4 times the value of the slope at an area of 1 square mile. On the Ontonagon Plain, however, the average slope at a stream length of 1 mile is 146 feet per mile, more than three times the slope in a 1 square mile drainage area. In the Ontona- gon area the channel slopes for any given stream length are much steeper than in the limestone and shale streams of the Shenandoah Valley and the profiles must also be correspondingly steeper, even though the area-slope relations are similar. This relation must be a conse- quence of the elongate character of the drainage basins. The relation of drainage area to length in an un- grooved is different from that in a grooved area. The normal relation found in the eastern United States (Hack, 1957, p. 63), and also exemplified by the upper Cranberry River (fig. 9), is L = 1.4 A”. In the grooved area the relation is approximately L=3.6A°~6. Thus, if a stream passes from an ungrooved area to a grooved area, there is a marked decrease in the rate of increase of discharge downstream, and this decrease causes the channel slope to approach zero at a decreased rate. SIZE OF BED MATERIAL The bed-material size was estimated by the measure- ment of 20 to 90 fragments selected randomly at most of the localities shown in table 4. No systematic rela- tion of bed-material size to channel slope or other fac- tors was found, although in the very small streams there is a very obvious increase of size with stream length. In streams having drainage areas larger than 2 square miles, no relation is apparent. The range in mean size of bed material of reaches in the larger streams is from about 12 to 125 mm. POSTGLACIAL DRAINAGE, STREAM GEOMETRY, ONTONAGON AREA, MICHIGAN B21 100 IIIII .... O IIIII CHANNEL SLOPE, IN FEET PER MILE IIII' |||||Il 1.0 I IIIIIII IlIIIII 0.1 1.0 10 100 1000 DRAINAGE AREA, IN SQUARE MILES FIGURE 13,—Relation of channel slope to drainage area in streams on the Ontonagon Plain, 0, channel in till and lake deposits; +, channel in bedrock; A, localities in Ontonagon River basin in which character of channel is not known. Some individual stream reaches are exceptions and show a systematic decrease in size of bed material. In both the Cranberry and Potato Rivers, directly below the bedrock reaches, the channels enter areas of till in which large boulders are uncommon. In these reaches both size and slope decrease systematically (p. B25, B29). In the area as a whole, however, the size of bed material is determined very largely by the material that encloses the channel and there is no apparent adjust- ment of slope in relation to size. The resistance of the bed and bank appears to be a more important factor controlling channel slope than does size of bed material. In the central Appalachians, channel slope is related in a rough way to the ratio of bed material to drainage area. N 0 such relation was observed in the Ontonagon area but, by comparison, the range in this ratio as well as the range in channel slopes is much smaller (fig. 14). CHANNEL CROSS SECTION Measurements of channel width and depth were col- lected at about 25 localities. The stream channels se- lected ranged in width from 6 to 143 feet (fig. 15) . A fairly good correlation was found between channel width and drainage area and the data compares closely with similar data obtained by Miller (1958, fig. 16) in New Mexico and with data obtained by Hack (1957, fig. 22) in Virginia and Maryland. The depth-width ratio on the Ontonagon Plain declines slightly with in- creasing drainage area and discharge as it does in other areas. The rate of increase of channel width is somewhat steeper in the Ontonagon area than in the central Ap- palachians. The curve of figure 15 has the equation W=9A°-6, where W is channel width in feet and A is area in square miles. The exponent for the Vir- ginia and Maryland streams is less than 0.5, as is the exponent suggested by Miller’s data for New Mexico 1000- A 3 /’z Lu — ’/ .4 -— d_/ ‘ r ’,,’ / E )— /‘_/ // E — // Pcfjtonjac , 0- asm / .— 100- [‘fi / I.|J : / / t3 - / ‘ / — Ontonagon / E — / area // L / ’/ E ' / 2/ o / ,x’ (j 10g // ’/’/ i f,/” z ... z _ < - 5 1,0 ..41.L. ........ ........ ........ 0.1 10 100 1000 RATIO OF MEDIAN SIZE OF BED MATERIAL, IN MILLIMETERS, TO DRAINAGE AREA, IN SQUARE MILES FIGURE 14.—Relation of channel slope to the ratio of bed- material size to drainage area. Area of scatter for locali- ties in the Ontonagon area compared with that for locali- ties in the Potomac River basin, Virginia and Maryland. B22 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY I I I I I I I l I I I I I I | I I I I I .50 _ _ O _ . _ O O O . O ' / o . / A . / O A// .10 - e .1 100 O ; : o o A // I; a / ° ‘ E I _ / O _ E p— A - g .05 — )K —50 k a //A Afi . I' ' A - I ,/ A p. E A // A Q Q — / o —- B / // x A // /A//A A A / .01 7/ 10 _ // _. _ // _ - // _ __ A/ A _ .005 — a 5 I l I I 1 l I I I l l | | I I | | I I 1.0 5 10 50 100 DRAINAGE AREA, A . IN SQUARE MILES FIGURE 15.——Re1ation of channel width and depth-Width ratio to drainage area in streams of the Ontonagon Plain. Equa- tion of line: W=9A°'“. perennial streams. The value of the exponent for the rivers of the Ontonagon area may be slightly higher because the drainage system is not in dynamic equilib- rium. The initial slope of the grooves was relatively steep and, because many streams still retain a steep slope, their profiles are less concave than is normal. The lack of curvature in the longitudinal profile is compensated by a greater rate of increase of channel Width. CRANBERRY RIVER The Cranberry River crosses the grooved Ontonagon Plain through a narrow corridor. The upper basin is nearly circular (fig. 6) and in the first 7 miles the drain- age area increases to 15 square miles. Below the junc- tion with Mason Creek, the largest tributary, the drain- age area increases only 4.5 square miles in the remaining 12 miles to the lake. The lower river is of particular O, depth—width ratio; A, channel width. value as an example for study not only because of the small increase in discharge but also because, although intrenched mostly in till, it crosses several bedrock thresholds and from sec. 3, T. 50 N., R. 41 W. through sec. 29, T. 51 N., R. 40 W. (fig. 16) it is deeply in- trenched in the Freda Sandstone. The bedrock has a strong influence on the behavior of the stream and changes the relative values of the variables that make up the channel equilibrium. The river was studied in two ways. The entire valley from Mason Creek to Lake Superior was traversed in the field. In addition, the lower course from sec. 29, T. 51 N., R. 40 W. to the lake was studied on a fully corrected optical model produced by using a Kelsh plot- ter, part of which is sketched in figure 16. The optical model permitted observations to be made of stream terraces that are difficult to measure in the field because POSTGLACIAL DRAINAGE, STREAM GEOMETRY, ONTONAGON AREA, MICHIGAN B23 Average channel slope, in feet per miIe 19 20 _. 7‘ '\ < \‘7 ”My/‘2???" k0 \ .~\.\\\\ 30 t 29 31 32 Z \\\\\\\fl AW“ // @1 s = elm‘ s A ‘\\‘ Area where valley bottom is in bedrock 14 ”>39? \\\\\\\\\ m, n ’l) ”1% (A) \\ fl/l \\ ////// Locality number 0 2000 4000 6000 FEET L__¢__4______4______J i:-,)\\\ ‘ // A \ \ // . \w Q3 0% l\\\ w“ MI, 22 \\\ yfl/ \ \l\\ ”WI 0 500 1000 FEET L______A__——_—J _ "Immlw 7, FIGURE 16.—Sketch map showing lower reaches of the Cranberry River. Numbers in circles are localities referred to in text; letters designate terraces. Sketch on left prepared by photo- grammetric methods using a Kelsh plotter. Sketch on right traced from aerial photographs; no horizontal correction. B24 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY of the dense cover of tall ferns and trees. Study of the model also permitted more measurements of channel slope than could be obtained from topographic maps. Above Mason Creek (fig. 16), the Cranberry River is intrenched about 60 feet in glacial till and lake sedi- ments. The till is stony and contains large erratic boulders. The streambed is floored with a lag concen- trate of rounded boulders washed from the till, many of which are 20 to 30 centimeters in diameter. The meander pattern consists of open smooth curves having a wavelength of about 800 feet. Below Mason Creek (loo. 2, fig. 16) , the stream passes across a resistant bed of felsite whose dip is about 30° to the north. For a distance of 2,500 feet, bedrock is almost continuously exposed in a channel only 30 feet wide within banks 3 to 4 feet high. Many large boul- ders are in the channel. Downstream where the felsite disappears beneath a gravelly clay till containing many stones the valley widens and a broad flood plain forms. The meander wavelength abruptly decreases to 350 feet, and the character of the streambed changes. In the bedrock reaches the bed is relatively flat and rocky and littered with angular boulders. There are few gravel bars. In the till reach downstream, the channel is floored with rounded cobbles, and there are many bars composed of fine gravel in and adjacent to the stream. Some of the bars are bare; others are covered with her- baceous plants which suggest that the bars are probably shifted by the annual spring floods and that the chan— nel is less stable here than in the bedrock reach. In this reach the bed material is finer grained than above Mason Creek, perhaps because it does not include a layer of bouldery till. At locality 3 (fig. 16) an outcrop of the Nonesuch Shale forms a threshold in the streambed and the chan- nel is again littered with boulders. At this point the valley walls are in a stony clay till about 70 feet thick. The channel abruptly straightens at the shale outcrop and the meander wavelength increases. Between localities 3 and 4 the character of the stream is rather monotonous. Depth of the valley averages 40 to 60 feet, mostly in till. The till is very stony in the lower part and contains large erratic boulders of igneous rock. The upper part is less stony and the top 20 feet is generally sandy clay. Outcrops of Freda Sandstone occur at several places in the channel shown as bedrock in figure 16. The outcrops, as well as the boulders in the till, provide a source of coarse bed material and the size of the material in the bed is maintained at a fairly high value (50—150 mm) all along this reach. Channel width is 35 to 40 feet. At locality 4, figure 16, the river becomes intrenched in Freda Sandstone and at locality 5 the top of the bed- rock is 40 feet above the stream. Twelve feet of silty clay till forms the top of the valley wall. The valley gradually narrows as it passes through this bedrock reach and at the narrowest point is only 300 feet wide. Bedrock outcrops in the channel are numerous and the bed material consists mostly of cobbles of Freda Sand- stone in both angular and rounded fragments. The channel slope steepens from an average of between 20 and 30 feet per mile to more than 50 feet per mile. The channel cross section is rectangular and shows few gravel or sand bars. The width averages about 50 feet. Mature trees grow on the banks. As shown in figure 16, the river has a tendency to form straight reaches between meander bends, as though the meander wavelength might be very large (as much as 1,500 ft) if the river were free to meander on an open- plain. At locality 6 the character of the bedrock changes from predominantly sandstone to shale or fine siltstone and the valley widens. The bed is mostly bedrock that forms long steplike sheets that slope gently downdip to the north. In many places cobbles, gravel, or other loose material are completely lacking on the streambed. The channel is unusually wide in this reach, in places as much as 125 feet. The meanders are again free and a few meanders of short wavelength are superimposed on the longer ones. At locality 7 the river is again in sandstone; intrench- ment of the stream is at a maximum, and the valley walls are about 80 feet high. Top of bedrock is 30 feet above the stream and the remainder of the valley wall consists of till and lake deposits. Average size of bed material at locality 7 is about 50 mm (fig. 17A). Downstream from locality 7 the bedrock surface slopes so steeply to the north that within only a short distance no more outcrops of bedrock are visible in the valley walls. The last outcrop in the channel is at locality 8. The north edge of the bedrock marks a profound change in the character of the valley as well as the channel. The valley widens from 500 feet to more than 1,600 feet, and a. flight of terraces borders the stream. At the point of emergence from the bedrock-walled valley or canyon (loo. 7), the meanders cease entirely but they begin again at locality 9. Be- tween these points, a distance of about 3,500 feet, the average size of the bed material is reduced from greater than 60 mm to less than 30 mm. At locality 7 the bed contains many angular boulders and cobbles of shale and sandstone. At locality 9 the material is mostly gravelly, poorly sorted, and rounded. Shale fragments are abundant but diminish in amount further down-stream until the bed material becomes a pebble gravel of rounded sandstone fragments. ' POSTGLACIAL DRAINAGE, STREAM GEOMETRY, ONTONAGON AREA, MICHIGAN B25 FIGURE 17.—Views of Cranberry and Potato Rivers: A, Oran» berry River near locality 7 in canyon cut in the Freda Sand- stone. B, Channel of Cranberry River in till near locality 12 showing boulder-y bed material of glacial erratics. 0, High bank cut by Cranberry River in the valley wall exposing the intermediate till and glacial erratics washed out of the lower layers. This locality is upstream from 17~B. D, Gravel bar typical of the small-scale meandering reach of the Potato River. B26 A marked change in the channel occurs in the vicinity of locality 9 similar to the change described between localities 2 and 3. The channel begins to meander but now has a much shorter wavelength than in the bedrock reach or in most of the upstream reaches. The average wavelength of the meanders is 450 feet, but these mean- ders are superimposed like a harmonic on larger mean— ders that have a wavelength of nearly 2,000 feet, slightly larger than the probable wavelength of the poorly de- veloped meanders in the bedrock reach. Whereas the channel above locality 9 is rectangular in cross section and contained in almost vertical banks, the channel be- low this point winds between rounded gravel bars. The width averages about 50 to 60 feet but in places is much larger. Banks are about 6 feet high but bars in the chan- nel, or adjacent to it are commonly almost as high as the banks. The average channel slope (measured along the stream and including the bends) also changes at locality 9. Above this point the slope is about 20 feet to the mile; downstream it flattens to about 10 feet to the mile—less than half as great. The valley walls in the reach between localities 7 and 10 are composed entirely of till and lake deposits. Al- though the lower layers of the till are quite stony, they contain no boulders and the river does not impinge against the valley wall. Therefore, the river does not pick up coarse rock fragments along the reach and the bed and bank materials are comminuted entirely from the material washed down from the bedrock canyon upstream. The change in character of the channel, as well as the presence of terraces, is probably associated with the ‘change in these materials as well as with a great increase in valley width. The changes are not unlike some of those described along streams in the Appala- chians that issue from the mountains onto a lowland plain underlain by relatively soft rock (Hack, 1957, p. 84; 1960, p. 91). Broad alluvial terraces form where the stream leaves the mountain. The channel slope sharply diminishes and the size of the bed material decreases. At locality 11, although the stream is still intrenched in till and the valley is fairly wide, the wavelength of the meanders again changes and becomes much larger. The change is associated with outcrops of extremely coarse bouldery till that appear in the lower valley walls at locality 11 and downstream. The till forms a lag con- centrate of coarse boulders which make the stream bed even rougher than in the bedrock reaches (fig. 17 B, C). The average size of the bed material is 125 mm, but the largest boulders are 700 to 800 mm (100. 1‘2). SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY, In spite of the change in size of bed material, the channel slope does not change appreciably; downstream from locality 10 it averages 10 to 12 feet per mile. The channel narrows, however, to about 40 feet and the banks are low (4 to 5 ft.). There are no sand or gravel bars in this reach. . Another change that begins at locality 11 is a narrow- ing of the floodplain which at locality 13 is only 130 feet wide. This change may not be associated with the change in bed material but may be associated with a terrace formed as a result of the backwater effect of a higher stage of Lake Superior, as is discussed on page B27. The terrace may be in an early stage of dissection and the narrow floodplain just developing. The alti— tude of the floodplain at this point is 625 feet, 23 feet above the level of the lake and only 9 feet above the main Lake Nipissing beach. Below locality 13 the elevation of the Ontonagon Plain declines and the valley walls become lower. At locality 14 the river crosses another buried ridge of bed- rock consisting of siltstone of the Freda Sandstone. This bedrock outcrop is significant because it indicates that the valley above this point cannot be underlain by any appreciable thickness of alluvial fill. The river must be degrading. Where it crosses the bedrock, the banks are 4 feet high and the channel widens to about 60 feet. The banks here are sandy. From locality 14 to 15 the valley walls and floor are till which, unlike the till upstream, contains no large boulders. The width of the channel again narrows to 1 less than 50 feet, and the size of the bed material declines to an average of about 12 mm; there are occasional riflies of coarser cobbles. Large boulders of glacial erratics are scattered on the bed. At about locality 15 the channel deepens and the river _ is affected by backwater from the lake. The bed is not visible below this point, and the river is bordered by a horizontal flood plain and terrace. The mouth of the Cranberry River is cut off from Lake Superior at low river stages by a gravel bar. The great power of the wave-generated currents of the lake relative to the power ‘ of the stream at ordinary flows is demonstrated by the almost straight shoreline that cuts across the narrow river estuary. ’ The Cranberry River suggests several conclusions re- lating to general problems of the morphology of river channels: ‘ 1. The channel slope is closely related to the nature of the valley walls and floor, and steep channel slopes are associated with resistant rocks in the bedrock reaches. POSTGLACIAL DRAINAGE, STREAM GEOMETRY, ONTONAGON AREA, MICHIGAN 2. Large changes in the size of bed material are asso- ciated with changes of slope in some places, as at locality 9; in others, the two changes seem to be independent of each other, as at locality 11 where the bed material changes but the slope remains almost the same. Probably size of bed material is only one of several changes that may accompany a change in channel slope, and its effect on slope may be balanced by other factors. 3. Channel width does not vary much along the river, but it seems to be greatest in the areas that have steep channel slope, as in the bedrock reaches. Thus the channel narrows somewhat at locality 9 where the bed material becomes finer. For some reason, at locality 11 it narrows even more in spite of the coarse bed material introduced at that point. 4. The narrow floodplain below locality 11 may be caus— ally related to the low slope. Perhaps high flood flows that are important in determining the chan- nel conditions pass through this narrow flood channel on gentle slopes with greater depth, and hence higher velocity, in spite of the greater channel roughness. 5. The most striking lesson to be learned from the Gran- berry River is perhaps the many changes that take place in the meander wavelengths, which range from 350 to 1,500 feet or more. The long wavelengths are associated with the bedrock reaches and with those that have coarse bouldery bed material. Short wavelengths occur where the bed material is fine and where the channel is bor- dered by bars that probably are disturbed by an- nual floods. The problem of the meanders is discussed on page B30. TERRACES OF THE LOWER CRANBERRY RIVER A vertical profile of the lower part of the Cranberry River valley was prepared by means of a Kelsh plotter from data obtained from aerial photographs (fig. 18). The area is heavily forested, but the aerial photographs used in making the map (fig. 16) and the profile (fig. 18) were taken in late April when the leaves were not on the trees. Even low terraces are generally visible on these photographs. Spot elevations, however, could be in error by several feet because of difficulties inherent in the photogrammetric technique. The terrace surfaces themselves are irregular and vary several feet in eleva- tion away from the stream. As a result, it should be borne in mind that there may be errors in the elevation and continuity of the various surfaces shown in figure 18. The following conclusions are drawn from study of the spot elevations and the terrace profiles. B27 1. Rather large variations are evident in the height of the channel banks. The banks are especially high in the upper bedrock canyon and diminish at the lower end of the canyon. 2. The terraces at the mouth of the bedrock canyon‘be- tween localities 6 and 9 indicate that the difference in elevation between the stream bed in the canyon reach and the stream bed in the till reach was formerly greater than now and that a rather broad transitional slope existed between the till reach and what is now the canyon reach. Presumably terraces c, d, and e are correlative, or nearly so, and are rock-defended terraces formed at a time when the river upstream flowed on the surface of the Freda Sandstone or in it at a shallow depth. At locality 11, however, the river was intrenched almost as deeply as it is now. Terraces a and b are perhaps also correlative and represent a some— what later stage of this adjustment. 3. The terraces h and j appear to be horizontal and presumably are graded to higher levels of Lake Su- perior. Terrace h must be graded to the level of Lake Nipissing, for it is exactly 16 feet above the present lake level and corresponds in elevation to the level of the well-defined nip formed by Lake Nipissing. Terrace j must correspond to some lower lake stage, but, upstream, appears to merge with the present floodplain. Neither of these base- level controlled terraces can be traced more than about a mile upstream from the lake. 4. The relation of the Nipissing terrace h to the rock- defended terrace d—e is noteworthy. The rock- defended terrace descends at a steep grade and, if the correlation is correct, at locality 11 is at a lower elevation above the river than is terrace h, the Nipissing terrace. This relation suggests that per- haps most of the canyon in the Freda Sandstone has been out since the high stage of Lake Nipissing, that is, in the last 3,600 years (Hough, 1958, p. 253), whereas the greatest amount of cutting in the till occurred prior to that time in the 4,000 years after the disappearance of Lake Duluth. The coarse bouldery bed material derived from erosion of the boulder clay at the base of the valley wall at locality 11 and below may have caused the rate of downcutting of this part of the river to be very slow. The boulders may form a limit to further flattening of the slope in this reach. Some evidence for this speculation is offered by the occurrences of bedrock above the main canyon reach (above loo. 4, fig. 16). At least 5 outcrops of bedrock are in the stream in this part of the valley but, except at locality 2 the river has not cut an appreciable depth B28 850’ 800’ ’7> 750’ 700’ 650’ SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY “Surface of Ontonagon Plain 600’ 0 2000 4000 6000 FEET FIGURE 18.—Profi1e of lower valley of Cranberry River from locality 5 to Lake Superior, showing profiles of upland, valley bottom, and terraces. on optical model. into them. Above locality 4 the surface of the bed— rock may inhibit downcutting until the longitu— dinal profile is steep enough to increase the ero- sional energy available. A similar condition may have prevailed below locality 4 when the stream flowed at the top of the Freda Sandstone now ex— posed in the canyon walls (locs. 5, 6) . POTATO RIVER The Potato River crosses a buried bedrock high area, and about 2 miles upstream from Lake Superior Where it leaves the bedrock it displays a change in the character of its channel even greater than that shown by the Gran- berry River. The lower course of the Potato River, shown in figure 19, is of interest particularly because of the sharp change in meander wavelength. At locality 1 and upstream the river is deeply incised in till and bedrock. The surface of the bedrock is 36 feet above the stream and is overlain by 45 feet of inter- bedded till and lake sediments. The till contains no large erratic boulders at this locality but there are some large boulders on the river bed that apparently have been washed down from upstream. As in the Cranberry River, the channel cross section is rectangular, banks are vertical, and gravel bars in the channel are small. Channel width is 60 feet and the depth 5 feet. Bed material consists of fragments of Freda Sandstone, shale, and erratic boulders. Average size is 68 mm but sorting is poor and 10 percent of the fragments are larger than 350 mm. One large boulder of diabase is 1,500 mm in diameter. The bed in most places is covered by cobbles, but in others sandstone bedrock is exposed over the entire channel. At locality 2 the bedrock surface slopes sharply down toward the north; the last outcrop is at locality 3 where the rock surface is exposed on the west bank 7 feet above the stream. Downstream from this locality there are no rock outcrops, and the locality marks a rather sudden change in the valley. Numbers in circles are localities shown in figure 16; FP, flood plain; a, terrace; ., spot elevation Data obtained by Kelsh plotter from aerial photographs. The meander wavelength decreases at locality 3 and remains small all the way to the lake. Large gravel bars, some of which are 5» feet high, mark the channel (fig. 25, lower cross section). The bed material aver- ages 12 mm or finer and there are no boulders. The largest fragments are about 7 5 mm. The low-water channel is narrow, averaging about 20 to 30 feet in width, but the entire channel cleared of trees may be as much as 125 feet wide. The channel is similar in character to that of the Cranberry River in the till reaches downstream from locality 2 and between 9 and 11, and is therefore characteristic of reaches in fine till in which the floodplain is very broad and in which all bed material must be brought from upstream. There is also a flattening of the slope, but it is not as pro- nounced as the change in the lower Cranberry River that takes place between localities 7 and 9 (fig. 16) . At locality 6 the river deepens and the channel nar? rows to 60 feet, a width that includes the bars. Bed material remains fine grained. At about this point on the river the backwater effect of Lake Superior changes the character of the channel and a short distance down- stream the meandering character ceases. This point is well Within the backwater zone of the lake and the channel is quite deep. The changes that take place as the Potato River leaves the bedrock canyon and enters the till area are summarized in table 3. Ranges are estimates. TABLE 3.—Oompartson of features of the channel of the Potato River in the bedrock canyon with those of the till area. downstream Feature Bedrock Till area canyon Average channel width ______________________________ feet" 60 60—125 Average size of bed material _________________________ mm" 60—80 10—20 Average channel slope ___________ feet per mile.. 22 12 Meander Wavelength ________________________________ feet" 800—1000 380 Sinuosity __________________________________________________ 1. 66 1. 57 POSTGLACIAL DRAINAGE, STREAM GEOMETRY, ONTONAGON AREA, MICHIGAN Terraces along the Potato River are shown in figure 19. The high terraces in the bedrock area correspond to the top of the bedrock surface and apparently are remnants of a broader valley that existed before the stream began to cut into bedrock. The terraces in the till reach downstream are apparently rock-defended terraces. No terraces border the river in the narrow valley just upstream from the lake and apparently no terraces can with certainty be said to have formed as a result of a higher base level. MEANDERS The example of the Ontonagon Plain should dispel any idea that meanders are phenomena of sluggish streams in areas that have been reduced to a state of low relief. The drainage here is youthful and the streams flow across the plain in almost canyonlike val— leys. Alluvial meanders, intrenched meanders, and compound meanders are found in the area and meanders occur in bedrock valleys as well as in till. As demon- strated by the Cranberry and Potato Rivers, the wave- lengths of the meanders change in response to changes in bed and bank materials; this change indicates that controls other than discharge affect wavelength. The extensive research that has been done on meandering streams is summarized in Leopold and Wolman (1960) and no attempt is made here to develop the theory of river meanders. Because of the complexity and variety of the meander patterns, however, the meanders of the Ontonagon Plain are worth considering because of the evidence they may add to knowledge of the problem. CONDITIONS THAT PRODUCE MEANDERS Leopold and Wolman (1957) have shown that mean- dering channels are end members of a continuum that extends from braided channels through straight chan- nels to meandering channels. In general, the channel form that develops is related only indirectly to load and is not necessarily a function of the amount of load carried by the stream. Data from many streams show that braided streams generally have a steep slope for a given discharge whereas meandering channels have a gentle slope for a given discharge. A critical line sepa— rates observed braided channels from meandering chan- nels on the basis of slope and discharge. The critical slope above which channels do not meander but are either straight or braided is defined by a line having the equation Sc=317 (2-0-44 where So is critical slope in feet per mile and Q is bank— full discharge in cubic feet per second (Leopold and Wolman, 1957, fig. 46) . B29 O 500 1000 (5) Locality number 1500 FEET . ”#an “minim” if; FIGURE 19.—Map of lower reaches of the Potato River. Numbers in circles are localities referred to in text. Prepared from aerial photographs using a Kelsh plotter. B30 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY Channels that have a small slope for a given dis- charge generally are in finer grained, less resistant ma- terials than channels with steep slopes (Hack, 1957; Brush, 1961) and it is clear that meandering streams must be associated not only with gentle slopes but with fine-grained material and also with relatively deep channel cross sections. Channels in the Ontonagon area meet these condi— tions. They are for the most part in till and lake sediments or in soft siltstone, shale, and fine sand- stone of the Freda Sandstone. The steepest initial slope possible on the Ontonagon Plain itself is only 50 feet per mile, the initial slope toward the lake. For small streams, this slope is probably far below the crit- ical value of Leopold and Wolman. Most, but not all, of the large streams have adjusted their profiles to a considerable degree and flow on much gentler slopes than the initial slope of the plain. Unfortunately it is not possible to compare the chan- nels of the Ontonagon area with Leopold and Wolman’s meandering channels because we do not have data on bankfull discharges in this area. There are, however, a number of straight river reaches in the area and all seem to be in areas where, for one reason or another, an initial steep slope has been imposed on a fairly large stream. Several examples can be seen on plate 1; the Iron River downstream from White Pine is one of the most striking—the slope there is quite steep for so large a stream and averages 24 feet per mile. In this region the cover of glacial drift is thin and the river flows in bedrock through the reach. Furthermore, it flows along the strike for much of the course, the strata are exposed in many places in the channel and there is lit- . tle fragmental material. In contrast, the Ontonagon River meanders in its lower course; it is cut in till, and its slope has been reduced to only 3.2 feet per mile, low for a stream of its size. Several other examples of straight reaches may be seen on plate 1. Both branches of the Union River have straight reaches as they flow off the northeast slope of the Porcupine Mountains. The Little Iron River has a straight channel west of Nonesuch where it flows on a steep bedrock slope. At a locality south of White Pine in sec. 16, T. 50 N ., R. 42 W. the Mineral River shows the change in pattern that accompanies a change in slope. Above the highway crossing, where the slope is 25 feet per mile, the stream meanders; below the highway, the stream crosses a buried bedrock high area, the slope increases to 100 feet per mile, and the channel straightens. The change in pattern can be seen in plate 1. Most of the streams in the Ontonagon Plain, even the smallest, have meandering channels. Two of the larger streams in the eastern part of the area, the Flint-‘ steel and F iresteel Rivers, are extraordinary because of the high sinuosity of their channels. They flow down low areas on the Ontonagon Plain in what are probably the axes of preglacial valleys. The initial slope before intrenchment was only about half as great as the slope west of the Ontonagon River where sinuosi- ties are not as high. In figure 20, the slopes of channels in the Ontonagon area are plotted against drainage area. The nonmean— dering channels have high-value slopes. MEANDER WAVELENGTH IN THE ONTONAGON AREA In the Ontonagon area the sharp changes in the wave— length of the meanders that accompany changes in the bed and bank materials are noteworthy. Meander wavelengths are related to the size of rivers. Leopold and Wolman (1960) show that in alluvial channels the: very conservative relation between channel width and wavelength is such that A=10.9W “’1, where A is meander wavelength and W is channel width and that this relation holds true for a very wide range of river widths. Dury (1962), in a study of misfit streams, measured the wavelengths of many meandering valleys as well as the alluvial meanders of the streams within them and shows that alluvial meanders in general have a wave- length (in feet) roughly equal to 90210-44 where A is drainage area in square miles. Valley meanders of the streams he studied had, on the average, 10 times the wavelength of the stream meanders. In the Ontonagon area the geology of the valley ob- viously is a complicating factor that affects the wave- length. A simple empirical classification is used to de— scribe the meanders of the area (fig. 21) . I ntre'rwhed meanders.—Intrenched meanders are in- cised in either bedrock or unconsolidated material. The depth of incision is greater than the normal depth of the stream channel. In effect, intrenched meanders are meandering valleys in which the stream meanders more or less in the same pattern as the valley, though the stream may be bordered by an alluvial plain. There are many examples of intrenched meanders in the Onto- nagon area, but perhaps the clearest are in the bedrock reaches of the. Cranberry and Potato Rivers, shown in figures 16 and 19. Some short reaches of the Ontona- gon River, west of Rockland are also examples (pl. 1). Alluvial meanders—The term “alluvial meander” is used to describe meanders of a stream developed in its own floodplain in which the depth of cutting is equal to the normal channel depth of the stream. Alluvial meanders are meanders of the stream channel rather POSTGLACIAL DRAINAGE, STREAM GEOMETRY, ONTONAGON AREA, MCHIGAN 1000_llll I I IrIIII I I I-1IIII I I Illlll I, II : u 500— - 1'1 3 A o: 100 A _ m _ ‘L — : I— “ o 0 Lu — O O _ Ed 50— o - Z - o o -I __ o E ._ O o 0 o o o O _. 9 o 0 c A U) _ o o O _ O E o o 10— c _ 5- _ — O _ _ 0/F|Intsteel RIver O .4 1IJII 1 llllll I lllllll I llllJII l I 1 5 10 50 100 500 1000 5000 DRAINAGE AREA, IN SQUARE MILES FIGURE 20.——Re1ation of channel slope to drainage area in meandering and nonmeandering streams in the Ontonag‘on area. than of the valley. The valley may be many times wider than the meander belt and the pattern of the val— ley is not necessarily related to the pattern of the stream. This kind of meander is the most common in the Ontonagon area; three examples are shown in figure 21. The lower course of the Ontonagon River has alluvial meanders (pl. 1). Alluvial meanders may occur within intrenched meanders as shown by the meanders of the Flintsteel River (fig. 21). Simple and compound meaMers.—Either class of meanders may be simple or compound. Simple mean- ders have one regular wavelength like the intrenched meanders of the Cranberry River or of the tributary of the Little Cranberry (fig. 21). Compound meanders have harmonics, that is, meanders superimposed on meanders of a longer wavelength, like the alluvial mean- ders of the Flintsteel River (fig. 21) ; the most striking example is the meandering reach of the Cranberry River at locality 10 (fig. 16). Data are from table 2. A, straight reaches; 0, meandering reaches. When the alluvial meanders are plotted, as on figure 22, they have a fairly conservative relation to the size of the drainage basin. A good fit is obtained with a line having the equation A=6OAVz (4) where A is meander wavelength in feet and A is area of the drainage basin in square miles. This line is close to, but not exactly the same as, Dury’s mean line for the wavelength of alluvial meanders. Two of the reaches with alluvial meanders have wavelengths far higher than the others. These localities are on the Cranberry River where the stream flows through coarse boulder till and the bed material is composed of large erratic boul— ders (loo. 1, 13, fig. 16). Intrenched meanders have higher wavelengths and a line having the equation A=22OAVI (5) passes through the field of points; thus, the intrenched B32 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY Tributary of Little Cranberry River Little Cran- berry River Potato River ALLUVIAL MEANDERS Cranberry River INTRENCHED MEANDER /' (Wm N Flintsteel River INTRENCHED MEANDERS AND ALLUVIAL MEANDERS 1000 l | 2000 FEET | FIGURE 21.——Typical kinds of meanders of streams of the Ontonagon Plain. meanders tend to have wavelengths about four times those of the alluvial meanders. The scatter is large, however, and the wavelengths of intrenched meanders of some smaller streams is at least five times larger than the wavelength predicted by equation 5. The Cranberry and Potato Rivers offer an oppor— tunity to examine changes in meander patterns in rela- tion to geologic conditions. As shown on figure 16, at- locality 1, the river has typical alluvial meanders, but they belong to the class of meanders with long wave- length (220A1/9). In this reach the river is flowing in coarse bouldery till and the bed is completely covered with boulders, many of them more than 250 mm in diam- eter. The river at locality 1 resembles closely the POSTGLACIAL DRAINAGE, STREAM GEOBJETRY, ONTONAGON AREA, MICHIGAN B33 10,000 “In: llllllll IlllF 5000 l lllll I AM\/ Illllll | lllll . W20 49” CF/ 1000 Illll MEANDER WAVELENGTH, A , IN FEET \h 0 % ° 0“ . /. lllll 500 A _ X - a ’ — . . 100 . _ : / — _ . u— _ / _ 50 - — I I | I | | I l l Ll I I I I I I I I I l I I I L41 I I I I I 0.5 I 5 10 50 100 500 1000 DRAINAGE AREA, A , IN SQUARE MILES FIGURE 22.——Relation of meander wavelengths in Ontonagon area to drainage basin areas. 0, alluvial meanders; A, large alluvial meanders of Cranberry River; x, intrenched meanders. reach at localities 11 to 13 where it flows in similar ma— terial and also has large wavelength. In the bedrock reaches between localities 4 and 7, the wavelength is also large, but the meanders are intrenched rather than allu- vial. The valley is narrow, and it meanders more or less in the same pattern as the stream. A third distinc- tive pattern occurs between localities 2 and 3 and between localities 9 and 11. In these reaches the valley is wide, the meanders are alluvial, and there are no bedrock out- crops; the bed material averages 25 mm in diameter or less and is brought down from the bedrock reaches above. Like the alluvial meanders of the Potato River, these meanders are bordered by high gravelly bars that are either bare or covered by grasses and herbaceous plants which suggest that the bars moved during spring flood. Compound meanders also occur on the Cranberry River. Between localities 6 and 7 the river flows through shaly beds in the Freda Sandstone and the valley widens. Although the meanders are distinctly of the intrenched type, the stream channel itself shows a tendency to meander in a smaller wavelength between low gravel bars. At locality 10, meanders of small wavelength (601115) are clearly superimposed on larger meanders (near 2201115), as shown in figure 16. Close examination of the channel, however, indicates that the stream has a tendency to develop another set of mean— ders, or bends, of even smaller wavelength. These are shown in figure 23. Although the small meanders are not complete, they do have a regular wavelength. They appear to be formed during a lower river stage than the peak spring floods that keep the main channel open. The Flintsteel and Firesteel Rivers have meandering patterns more complex than the others (pl. 1). They Flood plain Flood plaln 'l Channel 100 150 FEET Ian—«L—____l—l 0 50 FIGURE 23.—Detailed plan of short reach of Cranberry River, showing imperfect third order meanders. B34 both have long reaches in which compound alluvial meanders are well developed within the valley, and the valley itself has a fairly high sinuosity. Other mean— dering valleys wide enough to enclose alluvial meanders either are rare or have a very low sinuosity. The Flint- steel and Firesteel Rivers are different in that they do not flow in a grooved area but occupy valleys in which till is overlain by thick lake sediments. These rivers, therefore, were not constrained by grooves during the initial stages of valley development. Presumably the streams and the valleys eroded by them were free to meander in broader belts than other streams, and there- fore wide compound meanders could develop. In summary, the data in the Ontonagon area indicate that many alluvial meanders increase in wavelength in a regular manner with increasing drainage area and discharge. The rate of increase is not far from the rate reported by Dury (1960, fig. 2). Some alluvial mean— ders that wander freely across a broad flood plain have abnormally large wavelengths; these meanders are in channels lined with a lag concentrate of boulders. In— trenched meanders in bedrock, like those of the Cran- berry River, or intrenched meanders in till like those of the upper Iron River, also have larger wavelengths which are about four times the wavelength of the allu- vial meanders. Some streams display as many as three orders of meanders of different wavelength. MEANDER WAVELENGTHS IN THE SHENANDOAH VALLEY OF VIRGINIA The meander wavelengths of the Ontonagon area are similar to those of the Shenandoah Valley. The rela— tion of slope to drainage area also is similar, as are the channel cross sections. Furthermore, the river regimen is in some respects similar in that mean annual dis- charge increases with drainage area at the same rate in both areas. The Shenandoah Valley, like the On- tonagon Plain, contains both alluvial meanders and compound meanders, but the two do not generally occur in streams of the same size and compound meanders are not common. A typical stream in the limestone or shale lowlands of the Shenandoah Valley originates in a rather broad headwater valley with gentle slopes, and develops alluvial meanders along its channel within a distance of a mile or two from the head. Further downstream the channel becomes more rocky, the valley narrows, and at a river length of about 4 miles the enclosing valley narrows and begins to meander. In large rivers deep, well-developed intrenched meanders are the rule. The wavelengths of the Shenandoah Valley meanders are plotted in figure 24. They clearly are related to drainage area in the same manner as the meanders of SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY the Ontonagon Plain; hence the value 60A"é roughly describes the wavelength of the alluvial meanders and the value 220A“! fits the wavelength of the intrenched meanders. CONDITIONS CONTROLLING MEANDER WAVELENGTH Alluvial meanders formed inside meandering valleys, typified by the Flintsteel River (fig. 21), are common in many areas throughout the world, and a general theory has been proposed by Dury (1960) to explain them. Dury considers such streams misfit, that is, the valleys and the valley meanders are too large to have been formed by the streams that now occupy them. He believes that all such streams have valleys that were enlarged during a time when the climate was wetter and colder than now and that the dominant discharge was many times larger. The postulated discharge re— quired is enormous and is estimated at 80 times the pres- ent (Dury, 1960, p. 236). If greater discharge is the cause of valley meandering, then it obviously occurred rather recently, and Dury thinks it may have occurred in Atlantic or sub-Atlantic time, that is within the last 6,000 years. Dury (1962) has supported his theory by making borings in many meandering valleys and his findings show that the valleys commonly contain a fill of alluvium much deeper than the stream channel itself. This theory of misfit streams requires not only that the increase in effective discharge was very large, but also that it occurred within historic time and affected rivers in widely separated areas to a similar degree. These requirements are not easily visualized. The as- 10,000 5000 lvlllri. |III|IH IIIIIIII/lxllnr X .// | I IIIIH Illll 1000 500 \ I ll'lll 7i\ \c “5 74 {r . \ . l I Illlll 100: (J1 o IVII . ILIIIIII MEANDER WAVELENGTHv ix, IN FEET ._. O |llll 10,000 llllll I lllllll 10 I 100 1000 DRAINAGE AREA,A, IN SQUARE MILES ._. 0 FIGURE 24.—Relation of meander wavelength to drainage area for streams in the limestone and shale lowlands of the Shenandoah Valley. Data from aerial photographs, maps, and field measurements. The lines drawn through the fields of scatter are the lines of best fit determined in the Ontonagon area. X, intrenched meanders; O, alluvial meanders in upstream reaches. POSTGLACIAL DRAINAGE, STREAM GEOMETRY, ONTONAGON AREA, MICHIGAN sociation of the meanders with a deep fill in some valleys is, of course, not proof that the streams are misfit. It means only that aggradation has occurred. Application of the theory of misfit streams to the Ontonagon area shows that the entire Ontonagon Plain emerged less than 9,500 years ago and that only 4,200 years ago the Superior basin was occupied by Lake Nipissing (Hough, 1958, p. 253)- The lower valleys of streams like the Cranberry and Potato Rivers have been cut at least 15 feet since that time. The lower Cran— berry River contains good examples of large-wave- length alluvial meanders (200A1/2) that formed below the level of Lake Nipissing (fig. 17, locs. 12—14). If these meanders formed under conditions of greater dis— charge, the time must have been since 4,200 B.P. (2,200 BC), for at locality 14, more than a mile upstream from the lake shore, the elevation of the streambed is 613 feet, only 3 feet below the level of Lake Nipissing, and the entire flood plain downstream from locality 14 is at or below the Nipissing level. In the bedrock reach upstream from locality 7 the river is intrenched in till and bedrock in places to a depth of 80 feet. As shown by terrace profiles in figure 18, the lower 30 feet of this intrenchment was in bed— rock and must have occurred late in the history of the valley, presumably mostly in post—Nipissing time. In- asmuch as the bedrock part of the valley walls have the same meander pattern as the present stream, the mean- ders were formed either by the present stream or by a stream that meandered in a similar pattern. The con- clusion seems inescapable that the meander pattern and wavelength have not changed appreciably in the last 4,000 years and that the large-scale meanders have been formed within this period. The increase in the dominant or average discharge that would be necessary to produce the large-scale meanders of the Ontonagon area can be deduced from the data shown in figure 22. The curves show that the ratio of the wavelengths of large— to small-scale meanders is approximately 220 : 60 or 3.7 : 1, at any given drainage area. Conversely, the drainage areas at which large- and small-scale meanders have the same wave- length are related according to the ratio 13.5 :1. If it is assumed that the average annual discharge during the period of high precipitation postulated by Dury was proportional to drainage area, as is the present average discharge, then the average discharge must have been greater than that of today by a factor of 13.5 :1 and must have involved enormously greater precipitation. The average discharge would have been comparable in mag- nitude to the highest momentary discharges now re- corded in the region. B35 The large-scale alluvial meanders of the Cranberry River are especially difficult to explain as the result of increased average annual discharge, or of any change in climatic regimen. They involve the present channel and occur in a stream graded to the present Lake Su— perior; therefore, they are forming now or were formed very recently. The concept of the misfit stream as a general phenom- enon seems even less plausible when the meanders of the Shenandoah Valley are also considered. In that region the large-scale meanders are 3.5 to 4 times the length of the alluvial meanders and their length in feet is equal to 220AV2 (where A is measured in square miles) ; in other words, the intrenched meanders there have the same relation to the alluvial meanders of the valley as in the streams in the Ontonagon area. The meanders include those of the North Fork of the Shenandoah, which are in reality a meandering valley 80 feet deep in bedrock. It is not likely that this depth of cutting could have been accomplished in the last few thousand years. A more uniformitarian concept should be sought to explain the changes in the meander wavelengths of the Ontonagon area. That the wavelengths change where the bed and bank materials change is a coincidence that is repeated at several places in several rivers and that involves at least three kinds of material. This coinci- dence suggests that the cause is probably related to the change in material. No attempt will be made herein to formulate a complete explanation for the different scales of the meanders, but a partial explanation is suggested. In the Cranberry River, the essential difference be- tween the reaches that have large-scale meanders and those that have small-scale meanders is probably in the kind of material that the reaches 'are competent to handle at the dominant or effective discharge. There are three reaches in which large-scale meanders are well developed. The first is at locality 1 (fig. 16) where the bed and banks are mostly boulders, and where there are few gravel bars in the channel. Mature trees grow on the banks at the edge of the channel. In this reach, gravelly material is carried through, and out of the reach and the work done by the river is primarily that of scouring the bank and adding to the lag concentrate of boulders. Only a small proportion of the boulders are carried downstream because there are few of them in the channel below locality 2. Conditions are similar in the reach that runs from locality 11 to 14. Large scale meanders also occur in the bedrock reach between locali— ties 4 and 7. In this reach the bed is either composed of angular sandstone cobbles and boulders or is smooth bedrock. There are few gravelly bars, and mature trees grow on the steep channel banks. Probably the channel B36 Small-scale alluvial meander 0 10 20 3O 40 50 FEET L1_J_.;l—L_J APPROXIMATE SCALE m Bedrock Sand Sand, and cobbles FIGURE 25.——Generalized cross section comparing a typical stream channel in an alluvial reach with a stream in a bed- rock reach of the Potato River. and bed material in these reaches are adjusted to the highest discharges that occur only very rarely, and the annual floods modify the channel only slightly. The reaches of typical smaller scale alluvial meanders, however, as at localities 2—3 and 7—11, are quite different. In these reaches the banks are poorly defined. The river bed consists in part of a narrow low-water channel of bare rounded gravel moved from upstream. This chan— nel is bordered by wide gravel bars commonly covered by sparse grass and in places by dense growth of annual plants or low shrubs. The outer limits of the channel are defined, in places poorly, by a forest of mature trees. It is inferred from this distribution of the vegetation that the bed material and the material in the bars is probably moved around each spring, or at least every few years. The valley also differs in these reaches; it is much wider and commonly contains terraces as well as a Wide flood plain. In these reaches the channel must certainly migrate at a much faster rate than in the other reaches, for it has succeeded in shaping a much broader valley. Because the material handled by the stream is fine, more energy goes into shifting the channel and SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY handling material eroded from the steeper reaches up- stream. The difi'erences between the two kinds of reaches are illustrated diagrammatically in figure 25 taken from measurements of the channel of the Potato River. Figure 17D shows a typical gravel bar in the lower meandering reach of the Potato River. The difference in the character of the material handled by the stream probably produces a difference in the equilibrium of the channel and affects all the other vari- ables involved, including especially depth, width, and velocity, even though the discharge may be the same. Presumably the difference in equilibrium also includes the meander wavelength. Applying this concept to other areas we can expect, as we have found in the Shen- andoah Valley of Virginia, that meanders in streams that are eroding a bedrock channel will have larger meander wavelengths than streams that handle alluvial materials mostly transported from upstream. In the Shenandoah Valley, the small scale or alluvial meanders are confined to the headwater areas and the fine-grained alluvial material is derived by creep and wash from the residuum and soil formed on the valley sides (Hack, 1957, p. 85). SUMMARY AND CONCLUSIONS The foregoing study of the drainage system of the Ontonagon area was undertaken in order to observe a drainage network in a place where, because the character of the initial surface is known and dated, the evolution of the valleys that are now out into it may be inferred. The conclusions arrived at are now briefly recapitulated with little reference to the supporting data. THE DRAINAGE SYSTEM Development of the Ontonagon Plain—The Ontonagon Plain is a glacially grooved plain underlain mostly by clayey and silty sediments in part lacustrine, and in part glacial, of Valders and post-Valders age. The buried bedrock surface is irregular and overlain by as much as 200 feet of till, sublacustrine till, and lake sediments, predominantly fine-grained clay and silt. Of particular importance in the study of the stream valleys is the fact that the plain is cut by a set of discontinuous grooves that has exerted a strong control on the drainage pattern which formed as the plain emerged from glacial Lake Duluth. The lake withdrew from the grooved plain about 10,000 to 9,500 years ago. The evidence sug- gests that withdrawal of the lake was rapid, and that most of the plain was drained by 9,500 years B. P. At the time of Lake Nipissing, 4,200 years ago, the lake level was about 15 feet higher than present Lake Su- perior, and an escarpment was cut back into the till plain after that date. The base level of erosion in the POSTGLACIAL DRAINAGE, STREAM GEONIETRY, ONTONAGON AREA, MCHIGAN area has been at or below 616 feet ever since. The pre- sent base level (Lake Superior) is 602 feet. Uuttimg of the walleye. —Deep valleys have been cut in the Ontonagon Plain, penetrating both till and buried bedrock ridges. The valleys have been deepened by amounts roughly proportional to the quantity of water discharged through them. Small valleys have deepened little, even though they are at or close to the lake edge. The drainage network formed almost immediately as the lake withdrew. The drainage outline was deter- mined by the pattern of the grooves and other initial surface features of the plain and has changed very lit- tle. Piracies have occurred primarily as a consequence of downcutting as larger valleys grew and engulfed smaller ones. The break in slope at the nip formed by Lake Nipissing is preserved only in the profiles of the smaller valleys and shows as a convexity that is the more pronounced the smaller the valley. The character of the valley profiles is apparently determined more by the discharge and load of the streams than by changes in base level or, as Rubey (1952, p. 134) put it, by duties imposed from upstream. The headward migration of channels and divides seems to take place very slowly and concomitantly with downcutting of the valleys. The drainage geometry—Although attenuated be— cause of the grooves, the consequent drainage network, formed as glacial Lake Duluth receded, has many of the attributes of highly developed drainage systems in areas that have been erosionally graded. The charac— teristics of the drainage net on the Ontonagon Plain as defined by Horton (1945), including the rate at which the streams bifurcate as the drainage area in- creases, are similar to those of Appalachian drainage areas. The rate at which the drainage area increases with increasing stream length is also the same as in many other areas, although for any given stream length the drainage basin area is much less. The geometry of the drainage system is explained by the theory of maximum probability, that is, the system grew in a ran- dom manner as the lake receded from the plain and the runofl’ was discharged downslope through the grooves. The similarity to patterns in erosionally graded areas suggests that maximum probability largely controls the joining of streams in both environments. The import— ance of this generalization lies mainly in the ,conclusion that the rate of increase in discharge with stream length must, in most cases, be an independent variable in the channel equilibrium. The rate is determined partly by the climate of an area and partly by the probable six- tenths power rate of increase of length with drainage area. In a homogeneous landscape, the average rate of increase in drainage area relative to length is prob- ably not affected iby the amount of erosion in a given B37 area, the degree of gradation, the perfection of the ad— justment to geologic structure, or even by the average slope. However, local environmental factors such as structure obviously do affect the absolute value of the drainage area at a given stream length, and local changes in such factors must cause changes in the re- lation of area to length. Longitudinal stream pro/flea—‘The long profiles of the streams in the Ontonagon area are irregular as com- pared with streams in more completely graded land— scapes. The larger streams, hOWever, have some de— gree of concavity in their profiles. Because the profiles are adjusted to the nature of the materials that enclose the valley, stream reaches in bedrock are much steeper than reaches in till, other factors being equal. If the streams that are cut in till are considered by themselves, the relation of drainage area to slope is close and the gradients decrease systematically with increasing drain- age area and discharge. Because the shape of the drainage basin is an important factor in determining the discharge at any given point on a stream it also has an influence on the profile. In general, the profiles of the streams in the grooved part of Ontonagon Plain do not decrease in slope at the same rate downstream as streams in areas where drainage basin shapes are more normal. Meanders.—Meandering streams are the rule in the Ontonagon area. A few streams however, have straight courses and these streams invariably have un— usually steep slopes for a given discharge. They gen- erally have bedrock bottoms and banks. Streams such as the Flintsteel River that have very high sinuosity, on the other hand, are cut in lake sediments and have unusually low slopes. The relationship of slope to discharge that limits the meanders is similar to the re- lationship cited by Leopold and Wolman (1957). Meander wavelength in the Ontonagon area increases with discharge, but it also changes with the character of the bed and bank materials: in coarse materials the meander wavelengths at a given discharge are longer than in fine materials. Valley meanders also have a longer wavelength than the stream meanders. It is suggested that the meander wavelengths are determined by the discharge that is most effective in forming the channel in a given kind of material. No basis is found in this region to support the hypothesis of Dury (1960) that the large—scale meanders are inherited from an interval in late Pleistocene time when average dis- charges were larger than the present discharges. EVOLUTION OF THE LANDSCAPE The Ontonagon Plain is an erosionally graded land- scape in the sense that subaerial erosion has produced B38 some change in almost every part of it and the changes are related in an orderly way to environmental factors. On the other hand, the area as a whole cannot be con— sidered a single large system in dynamic equilibrium. Every reach of every stream might be regarded as a small system in equilibrium; that is, the form and slope are adjusted for the transportation of a certain load through a certain geologic structure under the avail- able regimen of discharge. A sudden change in the environmental conditions will produce a response such as a change in channel form. However, when a stream of some size such as the Little Cranberry River is con— sidered, it is evident that its valley and channel as a whole are not in equilibrium. The valley is constantly changing as erosion continues and the‘ changes involve the forms as well as the local relief. The Ontonagon Plain, therefore, differs from stream-eroded landscapes in areas such as the Appalachians where, presumably, all the elements of the landscape are mutually adjusted over large areas as the result of long—continued subaer~ ial erosion. At present the Ontonagon Plain contains many fea— tures, such as the grooves and shorelines, that are relict from past conditions, graded not by streams but by glacio—lacustrine processes. Eventually, all these fea- tures will be modified or obliterated by subaerial ero- sion. When all parts of the plain are mutually adjusted, equilibrium will be achieved throughout and there will be little further change in form except as the relief may slowly become lower. By making certain assumptions, we can infer what the evolution of the Ontonagon Plain has been in the past and how it will proceed in the future. If we as- sume that small and large valleys evolve in the same manner, we can then infer, for example, that the Gran- berry River valley once had the profile and other char- acteristics of adjoining smaller valleys and that the smaller valleys, as they enlarge and deepen, will evolve toward the form of the present Cranberry River valley and will develop the same slightly concave profiles. In general terms of evolutionary development, the typi— cal Valley of the Ontonagon Plain was initiated by the withdrawal of Lake Duluth. *The drainage basin area and discharge were fixed and determined by the initial features of the plain and have not appreciably changed. At first the valleys were shallow but gradually deepened at rates that were greater downstream. Because valley width was determined by the width of the meander belt, largely a function of discharge, the initial width of the valleys was not very different from the present. The ’ greatest change was in the profile. We know from study of the terraces that downcutting of the valley did not keep pace with the withdrawal of the lake. The SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY resulting break in slope was reduced both by» vertical doyvncutting of the stream and by a smoothing back of the break from the lake. The profile became first con- vex, then straight, then slightly concave. Concomitant- ly with‘the evolution of the profile, changes in channel shape took place. The initial channels, of the larger streams at least, must have been wider than they are now to compensate for the greater steepness. Initial channel slopes of some streams may have been too steep for the development of meanders, but as time went on meanders began to develop in streams, where they did not exist initially. There are few facts to guide any inferences about the evolution of interstream areas. Nevertheless, there has been some rounding and smoothing of the valley walls. Presumably the valleys widen somewhatthrough time as the meander belts themselves migrate from one side to the other of the valley. If base level remains the same indefinitely this widening may be accompanied by a rounding of the upper slopes and a softening of the sharp valley crests to smoother convex forms, until ultimately the valleys will be separated by rounded hills. If base level remains the same, the drainage pattern will retain its attenuated form for a great length of time, for the drainage basins can widen. only as the to- pography is lowered and as larger valleys engulf smaller ones. The development of the drainage to the present gives us no reason to believe that there will be anyrapid headward migration of divides away from the master streams. There is probably a limit to the distance that a divide can migrate without continued downwasting of the entire landscape. Eventually, the slopes in the area will be mutually adjusted and there will be no vestige of the grooved plain except an attenuated drain— age network. This evolutionary cycle need not be postulated in great detail; it has actually progressed to only a very limited extent. The remarkable aspect of the changes that have occurred thus far is the degree of gradation of the stream valleys in spite of what classical geomorphologists might call the extreme youthfulness of the topography.- The present develop— ments includes an adjustment of the profiles both to rock type and to discharge. The particular evolution that we envisage for the Ontonagon Plain, on the rather unlikely assumption that base level will not change, is in some respects simi- lar to the classic geographic cycle ofrDavis (1899) . The Ontonagon Plain is not presented, however, as a general example. The evolution that has cecurred there can exemplify only areas that have had a similar geologic history. Furthermore, the evolution is only superifi- cially like the theoretical cycle of Davis, and there are important differences. In the classic concept (Davis, POSTGLACIAL DRAINAGE, STREAM Gnom'rav, ONTONAGON AREA, MICHIGAN 1899), the evolution of the forms is dependent on base level. Although the larger streams on the Ontonagon Plain have indeed cut deeper valleys than the smaller ones, the gradation of none of them has much dependence on base level, except as base level limits the steepness of slopes in the area as a whole. The forms of the longi- tudinal profiles as well as the valley cross sections are dependent primarily on discharge and on what we might refer to as upstream factors. Even larger streams, such as the Cranberry River, except for a short backwater curve, enter Lake Superior at grades that are related to their increasing discharges; the efi'ect of an important recent change in base level is expressed only in a short stretch of terrace about a mile long. GRADED STREAMS The writer has suggested that the term “grade” or “graded” as applied to streams should be used simply to refer to the slope of any smooth adjusted channel (Hack, 1960, p. 83—85). The term is generally reserved for streams whose slopes are thought to be stable and de— termined by a load transported from upstream (Mackin, 1948). If the term is thus restricted, other channels such as those cut in hard rock are considered “un- graded,” even though they may have equally smooth and regular profiles determined by a load that is locally derived, or by other geologic factors. The Cranberry River well illustrates the point. This stream has a well- developed profile that may be divided into several smooth segments. It is, however, lowering its bed along most of its course, its profile is presumably chang- ing, and the bed load at any given place is composed partly of material transported from upstream and partly of material acquired locally from the bed or from the banks and valley walls. Nevertheless, the river is graded in the sense that its channel slope is adjusted in a manner similar to other streams in the same region for the erosion of a given type of material. The Cranberry River does have one or two short reaches that seem to meet the conditions suggested for the more special connotation of “graded” stream. The reach, about a mile long, between localities 7 and 10 (fig. 16) is an example. In this reach the bedrock floor of the channel is covered by cobbles and gravel, the material in the banks is entirely transported from upstream, the floodplain is bordered by terraces, and the size of the bed material decreases from an average of more than 50 mm to about 25 mm. At the upstream end of this reach the bed and bank material contains many angular fragments of shale as well as sandstone derived from immediately upstream. Within a short distance, how- ever, the shale, is comminuted and absent from the bed material; only the more resistant fragments are in the B39 deposits and they are more rounded than those up- stream. This condition ends near locality 10 where the stream picks up material eroded from the till that con- tains coarse cobbles. Below this point there is no further decrease in the size of the bed and bank materials. At locality 11 where the river encounters a coarse layer in the till the size increases again. The “graded” reach from localities 7 to 9 coincides with the place along the valley Where the river leaves its bedrock valley and enters a valley cut exclusively in glacial till. The profile changes from a steep one that is nearly without curvature and in which the average slope is more than 30 feet per mile, to a gentler profile that averages 16 feet per mile. It is a transitional reach between two profile segments in which the equilibrium conditions differ because of a difference in the enclosing materials. Within the transitional reach the profile must be more sharply concave than either up or down- stream in order to make the transition. Terraces exist because the bed material transported from upstream cannot be carried off in the gentler reach downstream and it is stored in the transitional reach until reduced by weathering and wear to smaller sizes. In all three reaches the river is presumably lowering its bed. This writer sees no advantage in calling the transitional reach “graded,” as though it were more stable, and the others “ungraded.” REFERENCES CITED Brush, L. M., Jr., 1961, Drainage ‘basins, channels, and flow characteristics of selected streams in central Pennsylvania : U.S. Geol. Survey Prof. Paper 282-F, p. 145—181. Butler, B. S. and Burbank, W. S., 1929, The copper deposits of Michigan: US. Geol. Survey Prof. Paper 144, 238 p. Davis, W. M., 1899, The geographical cycle: Geog. Jour., v. 14, p. 481—504. ” Dury, G. H., 1960, Misfit streams: problems in interpretation, discharge, ‘and distribution: Geog. Rem, v. 50, p. 219—242. 1962, Results of seismic exploration of meandering val- leys: Am. Jour. Sci., v. 260, p. 691—706. Dugmore, A. R., 1915, The romance of the beaver: Philadel— phia, J. B. Lippincott 00., 222 p. Flint, R. F., chm., and others, 1959, Glacial map of the United States east of the Rocky Mountains: New York, Geol. Soc. America, scale 1:750,000. Gravenor, C. P., and Meneley, W. A., 1958, Glacial flutings in central and northern Alberta : Am. J our. Sci., v. 256, p. 715- 728. Hack, J. T., 1957, Studies of longitudinal stream profiles in Virginia and Maryland: US. Geol. Survey Prof. Paper 294— B, p. 45—97. 1960, Interpretation of erosional topography in humid temperate regions (Bradley volume) : Am. Jour. Sci., v. 5, 258—A, p. 80—97. 1965, Geomorphology of the Shenandoah Valley, Vir- ginia and West Virginia, and origin of the residual ore deposits: U.S. Geol. Survey Prof. Paper 484. (In press.) B40 Hack, J. T., and Young, R. S., 1959, In‘trenched meanders of the North Fork of the Shenandoah River, Virginia: U.S. Geol. Survey Prof. Paper 354—A, p. 1—10. Hamblin, W. K., 1961, Paleogeographi'c evolution of the Lake Superior region from Late Keweenawan to Late Cambrian time: Geol. Soc. America Bull., v. 72, p. 1—18. Hamblin, W. K., and Homer, W. J., 1961, Sources of the Ke- weenawan conglomerates of northern Michigan: Jour. Geology, v. 69, p. 204—223. Horton, R. E., 1945, Erosional development of streams and their drainage basins; hydrophysical approach to quantitative morphology: Geol. :Soc. America Bull., v. 56, p. 275—370. Hough, J. L., 1958, Geology of the Great Lakes: Urban'a, 111., Illinois Univ. Press, 313‘ p. 1963, The prehistoric Great Lakes of North America: Am. Scientist, v. 5‘1, p. 84—109. Leopold, L. B., and Langbein, W. B., 1962, The concept of entropy in landscape evolution: U.S. Geol. Survey Prof. Paper 500—A, p. Al—AZO. Leopold, L. B., and Maddock, Thos. Jr., 1953, The hydraulic geometry of stream channels and some physi-ographic im- plications: U.S. Geol. Survey Prof. Paper 252, 57 p. Leopold, L. B., and Miller, J. P., 1956, Ephemeral streams— hydraulic factors and their relation to the drainage net: U.S. Geol. Survey Prof. Paper 282—A, p. 1—37. Leopold, L. B., and Wolman, M. G., 1957, River channel pat- terns: braided, meandering, and straight: U.S. Geol. Sur- vey Prof. Paper 282-B p. 39—85. SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY l Leopold, L. B., and Wolman, M. G., 1960, River meanders: Geol. Soc. America Bull., v. 71, p. 769—794. Leverett, Frank, 1928, Moraines and shorelines of the Lake Superior basin: U.S. Geol. Survey Prof. Paper §154—A, p. 1—72. ' Mackin, J. H., 1948, Concept of the graded stream: Geol. Soc. America Bull., v. 59, p. 463—512. ‘ Miller, J. P., 1958, High mountain streams :. effects of geology on channel characteristics and bed material: New Mexico Bur. Mines and Mineral Resources, Mem. 4, 53 p. . Rubey, W. W., 1952, Geology and mineral resources of the Hardin and Brussels quadrangles (in Illinois) : U.S. Geol. Survey Prof. Paper 218, 179 p. Snyder, W. M., 1962, Some possibilities for multivariate analysis in hydrologic studies: Jour. Geophys. Research, v. 67, p. 721-729. Strahler, A. M., 1957, Quantitative analysis of watersihed geo- morphology: Am. Geophys. Union Trans, v. 38, p. 913— 920. U.S. Department ‘of Agriculture, 1941, Climate and man, year- book of agriculture: Washington, U.S. Govt. Printing Oflice. U.S. Geological Survey, 1961, St. Lawrence River basin, pt. 4 of Surface Water Supply of the United States, 1960: U.S. Geol. ‘Survey Water—Supply Paper 1707, 437 p. Van Hise, C. R., and Leith, C. K., 1911, The geology of the Lake Superior region: U.S. Geol. Survey Mon. 52, 641 p. White, W. S., and Wright, J. 0., 1954, The White Pine copper deposit: Econ. Geology, v. 49, p. 675-716. U.S. GOVERNMENT PRINTING OFFICE : 1965 0—748—958 l lake 1a ty on stream lplSSlng f glac M70 1 o IFlOZ Oibmzciz 1y" :HKOZ wjmk tage of Leverett) N te stage of Lake Duluth In S Lake N APPROXIMATE MEAN DECLINATION, 1965 1a EXPLANATION 0 me o D High stage of Lake Duluth A Measurement local Trace of shorel Intermed (Lake Algonqu PROFESSIONAL PAPER 504—B PLATE 1 INTERIOR—GEOLOGICAL SURVEY‘ WASHINGTON. D C.—I9657664300 5 KILOMETERS 62 500 INTERVAL 20 FEET IS MEAN SEA LEVEL SCALE 1 CONTOU R DATUM SHOWING TRACES OF SHORELINES OF GLACIAL LAKES AND LOCALITIES WHERE MEASUREMENTS WERE MADE IN STREAM CHANNELS MICHIGAN, 9 MAP OF THE ONTONAGON AREA White / Survey topographic Matchwood 1 1949 1 Ontonagon, 1955; Greenland, Rockland ,and Bergland, 1956 UNITED STATES DEPARTMENT OF THE INTERIOR GEOLOGICAL SURVEY Base from us. Geological 1950 Pine quadrangles oqoflE—HMIIWH ' T, 9 j=. A f v SVSI (HIV HLI'IOLHH J0 V'EKVIG'IVO MPH ([NV'ISI ‘SISZNHSOHEHJ (INVfi T ‘ Y ‘ ' “ ‘ ‘4 V V‘ r 0—170 glam IfifiOIsseJoi'J 19mg [99130I599—0HVGT nudism ‘m €35 ’75 7DAY \"M/ \\ ' _ 49 Q 7 SEP ‘28 1965 \Q \ ' %eology and Petrogenesis of the Island Park Caldera of Rhyolite and Basalt Eastern Idaho GEOLOGICAL SURVEY PROFESSIONAL PAPER 504—C flangefi/ Geology and Petrogenesis of the Island Park Caldera of Rhyolite and Basalt Eastern Idaho By WARREN HAMILTON SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY GEOLOGICAL SURVEY PROFESSIONAL PAPER 504—C A stua’y of Me origin and occurrence of tfle products of eruption and collapse of a large magma c/zarnoer in waica liouiaI rflyolite overlay liouiaI oasalt UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1965 UNITED STATES DEPARTMENT OF THE INTERIOR STEWART L. UDALL, Secretary GEOLOGICAL SURVEY Thomas B. Nolan, Director For sale by the Superintendent of Documents, US. Government Printing Office Washington, D.C. 20402 CONTENTS Page Page Abstract ___________________________________________ Cl Petrology __________________________________________ 019 Introduction ....................................... 1 Petrography ................................... 20 Geology ........................................... 3 Rhyolite .................................. 20 The caldera ____________________________________ 4 Basalt _____________________________________ 20 Rhyolite of precaldera shield volcano .......... 4 Chemistry _____________________________________ 21 Caldera rim fault ........................... 6 Basalt _____________________________________ 22 Rhyolite domes on caldera rim _______________ 8 Latite _____________________________________ 24 Postcollapse rhyolite ash flows ________________ 10 Rhyolite ___________________________________ 24 Rhyollte lava flows 0f the eastern "m -------- 10 Interpretation of the Island Park caldera ______________ 25 Mafic lava flows on western part of precaldera . Ash flows ______________________________________ 25 shleld ___________________________________ 12 Ph . 1 hi t 25 Caldera fill _________________________________ 12 3'81“ F °ry ----------------------- 25 Interlayered basalt and rhyolite __________ 12 Petrogenesxs """"""""""""""""""" Basalt flows ____________________________ 13 Geologic and petrologic comparisons __________________ 26 Rhyolite domes _________________________ 13 Yellowstone Plateau ____________________________ 26 Rhyolite of the Yellowstone Plateau ______________ 14 Snake River Plain ______________________________ 27 Precaldera rhyolite ash flows _________________ 14 Columbia River Basalt __________________________ ' 28 Postcaldera “h flow ------------------------ 1'5 Bimodal volcanic provinces ...................... 28 Lava flows of Madison Plateau _______________ 15 Large calderas __________________________________ 29 7’ Basalt flows southeast of caldera __________________ 17 P _ f h b 11; h lit . t' 30 Basalt flows of snake River Plain _________________ 17 etrogenesus o t e asa -r yo e assoma Ion ------ .-—-—- surficial deposits ________________________________ 18 Relation of Island Park caldera to Snake River Plamuu 35 ., Age of volcanic rocks ____________________________ 19 References cited .................................... 36 III y ' }, r t ._) r» , f D :5 IV CONTENTS ILLUSTRATIONS Page PLATE 1. Reconnaissance geologic map of the Island Park caldera, eastern Idaho ________________________________ In pocket FIGURE 1. Index map of southern Idaho and adjacent areas _______________________________________________________ 02 2. Physiographic map of the Island Park region ___________________________________________________________ 3 3. Photograph of rhyolite tulf, of shield volcan0, exposed in cut 60 feet high __________________________________ 5 4. Northwestern part of Island Park caldera. Aerial photographic view west-northwestward from the center of the caldera _______________________________________________________________________________________ 6 5. Photograph of composite scarp at south edge of caldera _________________________________________________ 7 6. Photograph of southeastern part of Island Park caldera _________________________________________________ 9 7. Aerial photographic view northward along the front of the postcollapse rhyolite lava flow forming Black Mountain ________________________________________________________________________________________ 11 8. Photograph of obsidian agglomerate at top of rhyolite lava flow __________________________________________ 16 9. Aerial photographic view of vent zone in basalt of Snake River Plain ______________________________________ 17 10. Photomicrographs of rhyolite _________________________________________________________________________ 21 11. Photomicrographs of basalt __________________________________________________________________________ 22 12. Silica-variation diagram for rocks of Island Park caldera _________________________________________________ 24 13. Generalized silica-variation diagram for rhyolite and basalt of selected bimodal volcanic assemblages and for tholeiitic basalt and diabase_____.__.________________________.________T _____________________________ 31 14. Iron-magnesium variation diagram for basalt and rhyolite of Snake River—Yellowstone province and for Columbia River Basalt _____________________________________________________________________________________ 32 15. Diagrams of potassium-calcium, aluminum-calcium, and magnesium-iron ratios in basalt and rhyolite of Snake River—Yellowstone province and in Columbia River Basalt ____________________________________________ 33 TABLE TABLE 1. Chemical analyses of volcanic rocks from Island Park caldera and vicinity _________________________________ C23 ‘3 A4 ‘_ A. SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY GEOLOGY AND PETROGENESIS OF THE ISLAND PARK CALDERA OF RHYOLITE AND BASALT, EASTERN IDAHO By WARREN HAMILTON AB STRACT The Island Park caldera, in the northeastern part of the Snake River Plain, is an elliptical collapse structure 18 by 23 miles in diameter that was dropped from the center of a shield volcano composed of rhyolite ash flows. The western semicircle of the caldera margin is a single scarp in the northwest and a composite scarp in the southwest. Rhyolite domes and lava flows were extruded along the western rim during and after the period of collapse. The eastern semicircle of the caldera scarp has been covered completely by rhyolite ash flows, domes, and lava flows that were extruded along it. The caldera is filled, in upward succession, by rhyolite ash flows, interbedded rhyolite ash flows and olivine—basalt lava flows, and flows of olivine basalt alone. Rhyolite domes protrude through the basalt. The exposed rocks are middle( ?) and late Pleistocene in age. Basalt of the Snake River Plain, which laps onto the caldera shield from the west, was erupted from vents near the Island Park caldera during late Pleistocene and Recent time. The eastern part of the caldera is overlain by rhyolite ash flows and lava flows of late Pleistocene age from the Yellowstone Plateau. The rocks of the caldera are bimodal, consisting of uniform olivine basalt on the one hand and uniform highly silicic rhyolite on the other. The basalt is holocrystalline and diabasic and consists of labradorite, subcalcic augite, and magnesian olivine. The rhyolite is largely vitric; ash flows are mostly welded; and crystals of sanidine, high-quartz, and oligoclase are ubiquitous. A single flow of latite was found on the caldera rim. Eleven specimens were analyzed for both major and minor elements. The Island Park caldera is part of the Snake River—Yellow- stone province of intense Pliocene and Quaternary volcanism of olivine basalt and rhyolite. In this province, as in other bimodal volcanic provinces, rhyolite and basalt erupted from vents inter- spersed in both time and space, and simultaneous eruptions of both liquids from the same or nearby vents are known to have occurred. In the Island Park caldera, the eruptive sequence and geometry suggest that the large magma chamber contained liquid rhyolite overlying liquid olivine basalt. Several kinds of evidence indicate that the rhyolite of this and other bimodal volcanic provinces has formed by differentiation of basaltic magma. This differentiation cannot be explained in terms of fractional crystallization, but it can be explained in terms of an original tholeiitic basalt magma separating by liquid fractionation into rhyolite and olivine basaltic liquids in the proportion of about 1 to 5. Such fractionation may pos— sibly occur in the uppermost mantle or lower crust; there, rising tholeiite magma might split into immiscible phases, one rich in volatiles and fusibles (rhyolite) and the other rich in refrac- tories (olivine basalt), owing to instability of the initial homogeneous liquid caused by pressure decrease during ascent in a region of abnormally high thermal gradient. INTRODUCTION The Island Park caldera is perhaps the largest sym- metrical caldera yet studied anywhere in the world. Rhyolite was erupted early in the period of collapse, then basalt and rhyolite were erupted alternately from vents interspersed throughout the caldera floor, and fi- nally basalt was erupted alone. The caldera is so young that there has been little erosion of the tufl's, flows, ex- trusive domes, and fault scarps that comprise it, and there has been no apparent complication of the volcanic features by tectonism. The history of the growth of a broad shield volcano, the collapse of its central part, and the extrusion of magma during and after its collapse can be determined here with particular clarity. The caldera is significant petrologically because its rocks are a bimodal assemblage of uniform olivine ba- salt and equally uniform highly silicic rhyolite. These contrasting rock types apparently existed together as magmas—the rhyolite above the basalt—in the large chamber into which the caldera collapsed. The Island Park caldera lies in the northeastern part of the Snake River Plain of Idaho (fig. 1). The Yel- lowstone Plateau—the high northeast end of the Snake River Plain structural and volcanic province—extends to the east side of the caldera. The volcanic terrane rises gradually from an altitude of about 2,000 feet at the Oregon border to 8,000 feet on the Yellowstone Plateau. Flanking this terrane are still higher areas-— the Basin and Range province to the south, the central Idaho highlands to the north, and the high mountain blocks to the northeast. In the eastern half of the plain, volcanism has been concentrated along a northeast- trending axis, so that this part of the plain is higher in its center than along its edges; rivers flow near the Cl C2 116’" SHORTER CONTRIBUTIONS TO 114° GENERAL GEOLOGY 112° 110° l l I, Riggins O ,1 Z‘\/ \\‘ O Livingston / Salmon \ I 0 Council 44° —— 42° -_.l____ _____ IDAHO Dillon 0 MONTANA __ _ _ - _ — WYOMING Jackson 0 Pinedale Pocatello T_...__________—_._____ NEVADA I UTAH l O 50 100 150 200 MILES | l l l J FIGURE 1.—Southern Idaho and adjacent areas, showing relation of Island Park caldera to Snake River Plain and Yellowstone Plateau. margins rather than within the plain. The axial alti- tude of the Snake River Plain at the west side of the Island Park caldera is about 6,500 feet, 1,000 feet higher than the southern alluviated margin of the plain but only about 200 feet higher than the closer, northern margin. The Snake River Plain and the Yellowstone Plateau have been the site of intense bimodal basalt-and— rhyolite volcanism from Pliocene to Recent time. Near the Island Park caldera, the Snake Plain is bounded by the Centennial Mountains on the north and by the Teton Range on the southeast. Both mountain ranges are high young fault blocks that rise from the plain along long gentle dip slopes and face outward along precipitous fault scarps bordering the downdrop— ped blocks of Centennial Valley and Jackson Hole, re- spectively. Both mountain blocks have been much up- lifted during late Quaternary time and may be entirely Quaternary structural features. Rhyolite tufl' of early to late Pliocene age (Love, 1956) laps onto the Teton Range block and shares all or most of its tilting; simi- lar rhyolite laps onto the Centennial Mountains block. The rhyolite forms the outer part of the Yellowstone Plateau on all sides but the southwest (Boyd, 1961). It is buried by younger basalt and rhyolite in the central and southwestern parts of the Yellowstone Plateau and in most of the adjoining Snake River Plain. The high- lands flanking the Snake River Plain have been intensely block faulted during the Quaternary, yet the plain is virtually undeformed. ’ Blocks lessen in structural relief near the plain and disappear within a short distance of the edge of the volcanic terrane. Reconnaissance fieldwork for this report was done in July 1961. Most of the volcanic elements of this very young caldera are large, preserve their primar con- structional topography, and can easily be outlined on the basis of their topographic form. Most geologic units are large enough for depiction at a scale of 1: 250,- 000 (pl. 1). I mapped the caldera on aerial photo- graphs before seeing it in the field; so clear are the mor- phological elements that the photogeologic map l GEOLOGY AND PETROGENESIS, ISLAND PARK CALDERA, EASTERN IDAHO ‘ C3 (Hamilton, 1960a) was proved generally correct by the fieldwork. An abstract summarizing the field and petrologic features of the caldera has been published (Hamilton, 1962). That part of the caldera structure east of meridian 111° 15’ W. is shown on 1: 62,500 topo- graphic maps (Buffalo Lake quadrangle in the north, Warm River Butte quadrangle in the south), and many features of the rhyolite units can be recognized on these maps. The Island Park basin was recognized as a caldera by Steams, Bryan, and Crandall (1939, p.28), but they did not describe it. GEOLOGY The Island Park caldera is an elliptical collapse struc- ture 18 by 23 miles in diameter in the center of a rhyolite shield. The western semicircle of the scarp is exposed (fig. 2), but the eastern semicircle is buried beneath younger rhyolite. The interior of the caldera was flooded first by rhyolite, then by rhyolite and basalt, and finally by basalt alone, which erupted from vents scattered within the caldera. The contrasting magma types came from a single large magma chamber beneath the caldera. 10 MILES FIGURE 2.—Physiography of the Island Park region. The park, which is 18 miles across, is defined about its western half by the semicircular scarp of the Island Park caldera, and about its eastern half by the fronts of large rhyolite flows that erupted from the now buried scarp. The caldera rim is formed of welded rhyolite tuif, above which rise extrusive domes of rhyolite. The park is floored by basalt flows, and other basalt flows from outside the caldera lap onto the rim. Great flows of rhyolite, erupted from the crestal fissure on the Madison and Pitchstone Plateaus of the Yellow- stone Plateau, cover caldera extrusives in the northeast. quadrangle, U.S. Army Map Service, 1959. Photograph of pressed-relief edition of the Ashton 1 : 250,000 C4 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY The rim of the caldera is the remnant of a broad shield of rhyolite in which ash flows predominate over ash falls and lava flows and which is truncated by the caldera scarp. The present crest of the rim stands 1,200 feet above the plain to the south. The altitude of the crest is generally between 6,400 and 6,800 feet, but be- cause a younger basalt field laps onto the rim from the north, the western and northern slopes of the shield are much less exposed than is the southern slope. The caldera scarp is a single arcuate structural feature in the northwest, where it dwindles gradually northeastward from a maximum height of about 500 feet. The scarp in the southwest is a composite of several arcuate structural features which have an ag- gregate maximum height of 600 feet. Two or three rhyolite ash flows were erupted after the caldera collapsed, and these bury the caldera scarp in the north and in the southeast. The youngest of these ash flows is younger than all or most of the cal— dera fill and probably came from a vent near the caldera rim, but the other one or two ash flows came from vents in the Yellowstone Plateau. High, steep—sided domes of rhyolite were extruded in the west and south on the rim near the caldera scarp late in the period of collapse. The largest dome, Bishop Mountain, stands 1,100 feet above the surround- ing rim and was the source of a large lava flow of rhyolite that moved westward from the mountain. More extensive eruptions of viscous rhyolite along the caldera scarp produced thick lava flows and two or three low domes which completely cover the eastern semicircle of the caldera, but their vents define its position. These eastern flows are at least in part younger than the postcollapse ash flows. The western semicircle of the caldera rim is over- lapped from the outside by flows of basalt erupted from nearby vents. The eastern semicircle, after being buried by extrusions along its rim, was in part covered more deeply by very large lava flows of rhyolite erupted from a crestal fissure in the Madison Plateau part of the Yellowstone Plateau. THE CALDERA RHYOLITE 0F PRECALDERA SHIELD VOLCANO The shield volcano whose central part collapsed to form the caldera was a gently sloping circular cone. The shield is preserved as two arcs which together make semicircular Big Bend Ridge. The northwest arc is 9 miles long and 2 miles wide, and the southwest arc is 20 miles long and as much as 10 miles wide (pl. 1). The exposed rocks are rhyolite that dips radially outward at angles of a few degrees, parallel to the surface of the shield. Ash flows are much more abundant than are ash falls and lava flows. The best exposures of the shield are in the high cuts on the south slope along US. Highway 20 and 191 north of Ashton. The lowest unit exposed .in this section is a densely welded gray rhyolite ash-flow tuif (No. 1, fig. 3), and it is exposed almost continuously for 2.4 miles along the highway northward from Henrys Fork. No more than the upper 20 feet of the unit is exposed any— where, and the highway climbs updip parallel to the bedding. Discontinuously exposed above the ash flow is an irregular layer of loess, 0—8 feet thick, moved by slump and water. The loess is overlain by an unconsoli- dated rhyolite ash—fall tuft which in most places is air sorted, being composed of bedded crystal—rich tufl’ be- neath bedded pumice ash. The ash in many exposures is slumped, and its upper surface is irregularly eroded. The uppermost unit in highway cuts low on the shield is a soft, porous, orange-pink slightly welded rhyolite ash-flow tufl' that is locally agglomeratic; this unit thickens updip to more than 60 feet, so that the under- lying units disappear from view beneath it, and becomes a firmly welded pale-red lithic tufl'. Near the highest part of the shield along the highway, a well-bedded un— consolidated crystal-rich ash-fall tufi' overlies this pink tuff and is overlain in turn by as much as 50 feet of a slightly welded pale-red lithic ash-flow tuff. This up- permost unit also forms all the exposures along the highway in the two blocks dropped down between the three arcuate faults (pl. 1) that define the south rim of the caldera where crossed by the highway. The two welded ash flows that overlie the two un- consolidated ash falls are not present at the base of the shield, and the upper of the flows extends only a short distance from the rim. This relation is interpreted as due to limited deposition of the flows rather than to their erosion from the lower slopes. The ash flows ap- pear to feather out abruptly on the surface of the shield, and the lower limit of the upper ash flow is believed to be marked by a slight change in slope of the shield, which can be recognized on aerial photographs. ' The basal, densely welded ash flow in the highway section is probably visible in several other places on and near the south flank of the shield. West of the com- munity of Warm River and south of Snake River Butte, for instance, similar gray rhyolite is exposed discon- tinuously beneath a mantle of loess and till. Henrys Fork, where crossed by the bridge 2% miles west of Ash- ton, is incised 100 feet into what is probably the same ash flow. The upper part is dark—gray vitric tulf that is firmly welded but lacks secondary flow features; the middle and lower parts are light gray and devitrified and show irregular flow structure. The shield farther west is more complex, displaying _ rhyolite lava flows as well as ash flow tufl's. Two thick lobate rhyolite lava flows extend about 5 miles south- I GEOLOGY AND PETROGENESIS, ISLAND PARK CALDERA, EASTERN IDAHO C5 FIGURE 3.—Rhyolite tuft of shield volcano, exposed in cut 60 feet high. 1, Densely welded gray rhyolite tuft; 2, unconsoli- dated loess; 3: and 4,. unconsolidated bedded white tuft, sorted during airfall into quartz and sanidine crystals (3) and pumice (4); 5, soft, partially welded orange—pink rhyolite tuft that is agglomeratic at base. Roadcut along U.S. Highway 20 and 191, 0.8 mile north of Henrys Fork, 3 miles north of Ashton. westward from the vicinity of High Point. Although mantled by younger tufi, the form of the flows controls the surface topography. The lava flows are best ex— posed along the southwest edge of the shield, where their steep, eroded fronts are preserved. Each flow is at least 200 feet thick and consists of gray flow-contorted rhyolite containing blocks of spherulitic obsidian. The canyon of Blue Creek is incised along the valley formed between the sides of the adjacent lava flows. Normal faults having displacements of several hundred feet out this part of the shield and oifset the surface as well as the rocks. Rhyolite ash-flow tuffs underlie and overlie the lava flows, and one or two small thin lava flows, not distinguished from the tuff on plate 1, lie on the upper ash flow southeast of Blue Creek. The ash-flow tufi’s, which are poorly exposed, are mostly densely welded and gray. The ash-flow tuff of the shield is well exposed across the western caldera rim along a new forest-access road that runs westward to Antelope Flats from 761—374 H5—2 US. Highway 20 and 191 just south_of Swan Lake. A thickness of 150 feet within a single tufl' unit is ex- posed in high cuts where the road ascends the caldera rim. The ash flow consists of soft, pink and pale-red lithic rhyolite tuft and shows irregular flow structures which have a general dip of about 3° W.; this orienta- tion indicates that the tuff is part of the precollapse shield. A rhyolite complex unique among those seen is ex- posed in low outcrops 1 mile north-northeast of High Point. Squashed-vesicle welded tuif contains subordi- nate interlayers and lenses, each a few inches thick, of densely welded tufi'. Both types, which vary widely in texture and appearance, are unusual in that they lack phenocrysts. Strike and dip are variable even within small outcrops, but the dominant dips are moderately to steeply northward. Two tight folds having nearly horizontal axes were recognized. The folds show that, the tuff need not be a simple part of the tufl’ of the pre- collapse shield, for flow folds formed in the shield might . x 06 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY be expected to have axes parallel to the tangential strike of the shield—here north—northwestward—rather than eastward. Possible modes of origin include extrusion over a structural or erosional scarp during or after col— lapse of the caldera or upward frothy flow within a rhyolite vent. The northern part of the rhyolite shield was Visited only north of the Bishop Mountain cluster of rhyolite domes. Dense forest and debris from the domes cover this part of the shield, but its form shows that it is com- posed of rhyolite ash flows that clip radially outward. Photointerpretation indicates that one certain radial normal fault and a nearby probable one each offset the surface of the shield 100—150 feet. CALDERA RIM FAULT The western half of the Island Park caldera is defined by a semicircular zone of fault scarps 18 miles in di- ameter. The central part of the rhyolite shield vol- cano collapsed along these faults. A single fault scarp bounds the caldera on the northwest, whereas concentric scarps bound it on the southwest. The eastern half of the fault zone is buried beneath younger rhyolite, most of which was erupted along an arcuate zone that is believed to coincide approximately with the caldera margin. If this projection is correct, the caldera is elliptical and its major diameter is 23 miles long and oriented west-southwestward. The northwestern part of the exposed semicircle is a single curving scarp that rises from beneath the caldera fill at its northeast end to an even height of about 500 feet above the fill along its southwest half (figs. 2, 4). A rhyolite dome is offset by the fault near Green Can- yon Pass, but the ash flows are offset more; so the dome was extruded during caldera collapse. The east face of the high rhyolite dome, Bishop Mountain, also must be a fault scarp, because the crest of the dome is at the top of the face; had the fault scarp been present when the dome was extruded, the rhyolite would have flowed east- FIGURE 4.——Northwestern part of Island Park caldera. Aerial view in Winter, taken west-northwestward from the center of the caldera. . I i..— 7‘57? GEOLOGY AND PETROGENESIS, ISLAND PARK CALDERA, EASTERN IDAHO C7 ward down the scarp as well as westward down the rhyolite shield. The caldera scarp is largely buried by basalt flows in a 3—mile-wide gap east of Antelope Flat, and all but the crest of the dip slope of the rhyolite shield there is also buried by mafic lava. From Antelope Flat southeastward to the end-of the exposed rhyolite shield at Henrys Fork, the caldera boundary consists of concentric arcuate scarps in a zone Mg—Ql/g miles wide (fig. 5; pl. 1). Two faults are ex- posed in most of the zone, but three are present near the river. The faults branch and die out. Aggregate height of the scarps decreases from about 600 feet in the northwest to about 300 feet in the southeast. The same ash-flow tuff caps the crest of the main rhyolite shield \ and both of the lower blocks north of it along the high- way north from Ashton; thus the caldera is the result of collapse of the interior of the rhyolite shield rather than of explosive removal of the center of the shield. The scarps that mark the caldera-boundary faults are talus slopes that stand as steeply as 35°, although they generally are nearer 25°. The scale of the geologic map (pl. 1) requires that the faults be drawn at the foot of these slopes and that the talus not be shown as a unit distinct from rhyolite. The faults, however, are buried by the talus and are probably about vertical. As the scarps receded, talus buried their bases. The basalt flows that floor the caldera lap against and over the boundary-fault scarps (fig. 5) and do not seem to be offset by the faults. FIGURE 5.—Composite scarp at south edge of caldera. The rhyolite ash-flow shield in the middle distance, furrowed by radial gullies, dips gently south toward the snow-covered Snake River Plain. The inner (closer) of the two major caldera scarps visible in the middle distance dwindles to the right and is buried by postcollapse basalt flows near the highway. A rhyolite dome, Lookout Butte, projects through the basalt just to the right of Henrys Fork in the left foreground. CS SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY A rhyolite ash flow forms part of the south edge of the floor of the caldera. It dips very gently northward, opposite to the dip of the ash flows of the rhyolite shield. This rhyolite ash flow is considered part of the caldera fill, but it may actually be part of the rhyolite shield. More rhyolite is exposed beneath this ash flow in the gorge cut by Henrys Fork into the caldera floor, and it is not known whether any of this rhyolite is cor- relative with that of the shield. Total displacement on the caldera faults cannot be determined from the avail- able data. No scarps define the half of the caldera east of Henrys Fork. The northern caldera scarp dwindles gradually northeastward and vanishes beneath younger basalt, rhyolite tufi', and alluvium near the river. In the south the decrease in height east of the river is less regular, but the scarp also disappears beneath younger rocks. The lack of a scarp between the river and Moose Creek Butte might be due to cutting of a broad gap through the low rhyolite shield by Henrys Fork and the Warm River before the rivers cut their present gorges. The occurrence of river gravel (well exposed in a borrow pit by the old route of US. Highway 20 and 191) both on the divide between the present rivers and 400 feet above them east of Snake River Butte is consistent with this interpretation. The position of the eastern caldera boundary can be inferred from the vents of rhyolite lava flows and a dome erupted along an arcuate zone that continues the trend of the exposed western semicircle. As rhyolite domes were extruded along the western semicircle, it is reason— able to assume that rhyolite domes and flows were erupted also along the eastern semicircle. If this is correct, the eastern flows were the more voluminous, for they completely bury the scarp. The vents that define the eastern semicircle are noted here in counterclock- wise order. Warm River Butte is a rhyolite dome that is elongate east-northeastward. Moose Creek Butte is a steep—fronted lava flow (or broad dome) whose crest is on the same trend. High points along young rhyolite lava flows define other vents on the arcuate trend in the Snow Creek Butte-Buffalo Lake area of the east mar- gin of the structure. On the northeastern perimeter, no vents were recognized for a distance of 15 miles, and the position of that part of the caldera boundary as shown on the map is inferred. Because the exposed scarps decrease in height east- ward in both the north and the south, it may be that the central collapse area had the geometry of a trap door, hinged in the east and dropped farthest in the west. If so, a scarp never existed in much or most of the in- ferred eastern half of the structural feature. This seems less likely than does the possibility that the eastern scarp has been buried by younger rhyolite. Even though the exposed scarps have an aggregate height of only about 300 feet near their southeastern end, the postcollapse caldera fill in the nearby gorge of Henrys Fork is at least 300 feet thick, so that the minimum olf- set along the caldera fault near Henrys Fork is 600 feet. It is unreasonable to infer disappearance of this dis- placement within 2 or 3 miles to the east along a struc- tural feature that is so regular throughout its long ex- posed part. The altitude of the caldera rim in the western semi- circle increases from about 6,100 to 6,300 feet at its east ends to 6,500 to 6,850 feet in its western part. This is remarkably little variation, if one considers the size of the feature. The central part of the precaldera rhyolite shield collapsed very evenly. RHYOLITE DOMES ON CALDERA RIM Steep domes of rhyolite were erupted along the cal- dera scarp during collapse of the structural feature. Largest and highest of these domes is Bishop Mountain, in the west; other domes occur near Bishop Mountain, and still others form Snake River Butte and Warm River Butte, in the south. Constructional slopes pre- served on the sides of the domes are as steep as 35°, and steeper slopes may have formed originally. The distinction between rhyolite lava flows and domes is arbitrary. Both spread laterally from their ents, and both generally have steep fronts. Domes have greater ratios of height to area, but they need be no higher than the flows are thick. The domes are of various ages. Bishop Mountain and Snake River Butte are broken by the faults of the caldera rim and are older than all or much of the col- lapse. The unnamed dome adjacent to Bishop Moun- tain on the east is only slightly offset by one of the faults and is therefore older than only the latest part of the collapse. The two low domes southeast of this dome are entirely within the caldera, so that their age relative to that of collapse is unknown. Warm River Butte dome rises above tufl’ which hides the scarp, and the dome cannot be dated relative to the faulting. Bishop Mountain (fig. 4) stands 1,600 feet above the broad caldera floor and 1,000 feet above the caldera rim. In the east the dome is broken by the caldera scarp; in the west the dome becomes a rhyolite lava flow that covers nearly 25 square miles to a depth of as much as 1,000 feet. Light-gray richly spherulitic lithoidal rhyolite containing sparse crystals dominates the sur- face rubble of both flow and dome. Blocks of spheru- litic black obsidian are abundant on the crests of both. Because most very young rhyolite lava flows in the region are capped by unconsolidated agglomerate con- taining abundant blocks of similar obsidian, it is likely that this material once covered the entire flow but has GEOLOGY AND PETROGENESIS, ISLAND PARK CALDERA, EASTERN IDAHO since been largely eroded from its slopes. The Bishop Mountain flow diverted Henrys Fork from a previous course along the north base of the rhyolite shield volcano to its present course southward across the caldera. Island Park Reservoir (pl. 1; fig. 2) broadens west-southwestward, away from the dam at its east end; the river previously flowed west-south— westward also, but it was dammed by the rhyolite flow and diverted into the caldera. Still younger basalt buries the old channel farther west. The old course of the river followed the broad constructional valley be- tween the foot of the rhyolite shield and the rhyolite ash flows and basalt to the north, The small, unnamed dome northeast of Bishop Mountain was also extruded from a vent on or very close to the caldera fault, and it stands 600 feet above the adjacent caldera rim. Green Canyon Pass is in the valley formed by the opposed fronts of the small dome and the Bishop Mountain dome. The small dome is crossed by the caldera scarp, and the part of it on the caldera side dropped about 200 feet. The caldera scarp where it cuts the adjacent tuff shield is 500 feet high, so the downdropped part of the dome has been offset less than has the shield. The dome is considered to have been extruded along the caldera fault during col- lapse. Poor outcrops near Green Canyon Pass show FIGURE 6.—Southeastern part of Island Park caldera. 09 that both the small dome and the Bishop Mountain dome consist of light—gray finely porphyritic lithoidal rhyolite. Rubble of this material covers the slopes of the shield between these domes and Island Park Reservoir. Another rhyolite dome of the Bishop Mountain clus- ter has an area of about 1 square mile and a height of 300 feet and stands entirely within the caldera east of Bishop Mountain. The rubble on this dome is light- gray sparsely porphyritic spherulitic rhyolite. The extrusive rhyolite forming Snake River Butte dome rises 400 feet above the caldera rim in the south and is exposed in clifl's above Henrys Fork for a depth of several hundred feet beneath the rim level. The rhy- olite in the cliffs is white and massive, but the blufl's at the southwest front of the dome expose richly pheno- crystic brown rhyolite. The dome appears to be broken along its north side by a caldera-boundary fault (pl. 1; fig. 6, right edge) ; the adjacent part of the downdropped block of rhyolite was not visited, but a semicircular high area upon it is interpreted on the geologic map (pl. 1) to be the downdropped part of the same dome. A small, unnamed dome adjoins Snake River Butte on the west. Warm River Butte dome rises 500 feet above flanking ash flows east of Henrys Fork (fig. 6). The dome is on the trend of the arc of the caldera rim and presumably The extrusive rhyolite dome, Warm River Butte (center), projects through an ash flow that slopes gently to the right (west); the steepfronted rhyolite lava flow, Moose Creek Butte, is at far left. the caldera floor (plain, crossed by railroad). center distance. The ash flow ends at an irregular primary front at the river canyon, toward which slopes the basalt of Aerial view southward. Teton Range on left skyline; Teton Valley in 010 I SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY was extruded along or near the rim fault. The eroded sides of the butte consist of uniform light-gray rhyolite containing abundant phenocrysts. Obscure flow layers on the south side strike parallel to the hillside and dip vertically or steeply into the hill. Some of the rhyolite encloses granules of dark glass. Much of the rhyolite in the crest of the dome is richly spheroidal and is ap- parently devitrified glass. Similar material has prob- ably been eroded from the steep sides of the dome. Warm River Butte is elongate east—northeastward along the presumed trend of the caldera fault. A little dome only 100 feet high lies along the axial trend of the butte at its northeast base. POSTCOLLAPSE RHYOLITE ASH FLOWS The oldest rocks in the caldera margin that are cer- tainly younger than the collapse are welded rhyolite ash—flow tuifs which crop out in the northern and south— eastern parts (pl. 1). Surfaces of the thick ash flows slope gently toward the caldera and are considerably dissected. The caldera scarps trend toward the ash flows but do not displace them; the ash flows or unknown units buried by them have filled smoothly across the scarps. Two of the ash flows were probably erupted from a vent along or near the eastern part of the caldera fault zone and are described here; a third probably came from the Yellowstone Plateau northeast of the caldera and is discussed subsequently. The ash-flow tufl' bounding Island Park on the east side from Eccles Butte to Warm River Butte ends in an irregular blufl" (fig. 6) that is probably the original front of the flow or at least of its welded part. Basalt flows correlative with those of the caldera floor cover the tufl' on the rim of the gorge of the Warm River where the river leaves the caldera at the south. The youngest basalt flows of the floor cover the front of the ash flow in the south, but the exposed front of the rest of the ash flow indicates it to be younger than most of the caldera fill. The Warm River runs in and has slightly deepened the depression between the rhyolite front and the basalt sloping gently toward it. The ash flow is poorly exposed. Where traversed along Fish Creek and its north fork, the sheet consists of densely to slightly welded gray and light—purplish- gray porphyritic rhyolite that is generally massive but locally shows flow structures that are mostly subhori- zontal. Only a single ash flow may be present. The Warm River Butte rhyolite dome (fig. 6) projects through the tuff, which laps on to the lower slopes of the dome; whether the dome was extruded through the tuff or the ash flow was erupted around the dome is not known. The ash-flow tufl' east of Moose Creek Butte is similar lithologically to that around Warm River Butte and is probably part of the same unit. At its northern ex- posed limit, this eastern part of the ash flow consists of moderately welded light-gray tufi' that contains frag- ments of stony rhyolite and is underlain by the densely welded purplish-gray tufi' of the interior of the flow. The ash flow overlies the basalt flows exposed south- east of Moose Creek Butte and is in turn overlain by rhyolite lava flows that came from the northeast. The lava flow forming Moose Creek Butte probably also overlies the ash flow. The surface of the ash-flow sheet slopes gently west- ward between Eccles Butte and Flat Canyon, west- southwestward between Flat Canyon and Warm River Butte, and southwestward south of Warm River Butte. East of Moose Creek Butte the surface slope changes gradually from southwestward to southward as the sheet rises toward Snow Creek Butte. These slopes converge upward toward Snow Creek Butte; so this area is presumed to contain the vent from ‘which the ash flow was erupted. The vent apparently lay along or near the buried caldera fault zone. Any structural feature that formed when the collapse occurred and the ash flow was erupted has been overfilled by younger rhyolite lava flows. The abrupt front of the ash flow (fig. 6)1 indicates that the tuff solidified from a dense mass that behaved mechanically like a flow rather than like a clOud. The effective viscosity of this flow was much less than that of the rhyolite lava flows described later, for its upper surface is smooth and its area is great. Unwelded tufi' may have been eroded from the ash-flow front to expose the front of the welded part of the flow. The northern part of the caldera margin in the Buf- falo River area is covered by ash—flow tufl'. Two sheets are present: a densely welded tufi' and, overlying it in the east, a slightly welded tufl'. The densely welded sheet, whose surface slopes southwestward, probably erupted within the Yellowstone Plateau; it is described subsequently. The slightly welded sheet, which consists of soft, gray to orange-pink rhyolite tufl’ and whose dis- sected surface slopes westward, presumably erupted near the caldera rim. RHYOLIT’E LAVA FLOWS OF THE EASTERN RIM Large lava flows of viscous rhyolite that were ex- truded from vents near the presumably buried eastern rim of the Island Park caldera form a plateau which is about 15 miles long from north to south and 10 miles wide. Rhyolite domes were extruded around the ex- posed western half of the caldera rim. The flows of the eastern group apparently represent more voluminous eruptions of slightly less viscous rhyolite than that of the domes; flows and domes bear a similar relation to the caldera structural feature. VV “.7." Y rv_l GEOLOGY AND PETROGENESIS, ISLAND PARK CALDERA, EASTERN IDAHO Cl]. Individual flows have generally steep fronts and sides, and their boundaries, where recognized, are shown on plate 1. About half the rhyolite pile is covered by What appears to be a single rhyolite flow whose high points and presumed vents are along an arcuate zone in the Buffalo Lake—Snow Creek Butte area. The largest lobe of the flow moved west—northwestward, formed Black Mountain, and reached the caldera floor; smaller lobes moved southward and northward. ‘The flow is heavily forested, but concentric flow ridges are obvious on its little~eroded surface (fig. 7). Next oldest among the rhyolite lava flows is the south- eastern one, which flowed southward from vents now buried. Although this flow has been much more eroded than has the Black Mountain—Buffalo Lake flow, its primary lobate form is well preserved. The nearly round Moose Creek Butte flow, which is isolated from the others of the group, shows similar preservation of form and is perhaps of about the same age as the south- eastern flow. The front of the youngest rhyolite lava flow (that forming Black Mountain) on the eastern rim is par- ticularly high (fig. 7), and its height remains unchanged as the flow is traced from the certainly older ash flow FIGURE 7.—-View northward along the front of the postcollapse rhyolite lava flow forming Black Mountain. Concentric flow ridges cross the flow above its steep primary front. The snow-covered and largely treeless plain is formed on alluvial obsidian sand. 012 SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY of the eastern rim to the contact with basalt and allu- vium. The flow must overlie all or most Of the obsidian sand and basalt flows Of the adjacent caldera fill. MAFIC LAVA FLOWS 0N WESTERN PART OF PREC’ALDERA SHIELD Antelope Flat, on the western rim of the Island Park caldera, is a rolling plain formed Of mafic lava largely covered by a thick deposit of loess. The topographic crest Of this lava plain is within Antelope Flat several miles west Of the caldera scarp. Presumably this crest is the vent area, although the topography is complicated by two long, low fault scarps striking west-northwest- ward (pl. 1). Cinders mark another vent north of the center of Antelope Flat, but any initial conicalform of the cinder accumulation has been destroyed by erosion. High Point, at the southeast corner of Antelope Flat, is a large and considerably eroded mafic cinder cone. The only specimen collected from the lavas of Antelope Flat is an Odd mafic latite. The mafic lava flows overlie the rhyolite ash flows Of the precaldera shield volcano, but their relation to the large rhyolite lava flow forming Bishop Mountain was not determined. CALDERA FILL The volcanic rocks filling the Island Park caldera pre- sent a surface of Olivine-basalt flows (figs. 4, 6), erupted from vents scattered within the caldera, through which project extrusive domes of rhyolite (figs. 4, 5). A sec- tion 500 feet thick is exposed in the gorge where Henrys Forks leaves the caldera. Rhyolite ash flows are at the bottom of the section, interlayered rhyolite ash flows and Olivine-basalt lava flows are in the middle, and ba- salt flows are only near the top. The rest of the caldera floor is not eroded. The Older flows of the group are much eroded, and their fronts are greatly modified. The northernmost flow, here grouped with the flows of the eastern rim although it may have come from farther east in the Madison Plateau, lies on the rhyolite ash flow Of the eastern rim and has been incised deeply by North Fork Split Creek. The two or more southern flows, crossed by Flat Canyon, are even more eroded. The uppermost southern flow may have been extruded in part from a vent near the highest preserved point on the flow, at an altitude of 7 ,490 feet, and it overlies and is thus younger than the ash flow Of the eastern rim. Presumably the flow beneath the uppermost southern flow is also younger than the ash flow, because the surface of the ash flow is uninterrupted on both sides of that flow. The rhyolite lava flows are very similar morphologi- cally to those Of the Yellowstone Plateau, which are de— scribed subsequently. They have an upper zone of Obsidian agglomerate consisting of unsorted blocks of granular porphyritic black obsidian in an unconsoli- dated glass—shard matrix. Beneath the obsidian Zone, as exposed along Split Creek, is a zone at least 300 feet thick of flow-contorted and chaotically interlayered glassy and lithoidal rhyolite. Analogy with the flows Of the Yellowstone Plateau indicates that at a greater depth the flows are probably composed of wholly lithoidal rhyolite. The Obsidian of the uppermost southern flow out by Flat Canyon is largely devitrified, a state presumably indicative of the relative age of the obsidian. The still Older lower flow was not visited. Moose Creek Butte was visited only at its southeast slope, where massive gray lithoidal rhyolite forms large rounded outcrops. Sparse blocks of welded rhyolite tuff and of basalt project through the thick loess and soil covering the up- permost southern flow where the flow was studied near North Fork Fish Creek. No exposures wer ‘ seen which demonstrated whether these blocks are him rhyolite ash flows and basalt lava flows which overlie he rhyolite lava flow or whether they are glacial erratics, but the latter possibility seems more likely. INTERLAYERED BASALT AND BHYOLITE At Upper Mesa Falls, the gorge of Henrys Fork is 400 feet deep. The lowest “cooling unit” (Smith, 1960) of welded rhyolite ash flows has an exposed thickness of 125 feet, but the base of the unit is covered. Partings separate successive ash flows within this cooling unit, but columnar joints are continuous across the partings. On the east side of the river, an Olivine—basalt flow over- lies these ash flows and marks the base of a poorly ex- posed section about 200 feet thick Of interbedded lava flows Of olivine basalt and welded ash flows of rhyolite. This same section is exposed in clifi's on the inaccessible, west side of the river but, there, appears to consist of rhyolite ash flows only. The basal cooling unit, how- ever, is exposed continuously across the river, and the reason for this apparent discrepancy between higher sections is not known. The upper 75 feet of the section on the east side of the river is made up of basalt flows alone; this interval was not seen on the west side. Lower Mesa Falls are cut in a welded rhyolite ash flow which has an exposed thickness of 60 feet. Because units were not traced along the gorge, the relation of this ash flow to those at Upper Mesa Falls is unknown. Interbedded rhyolite and basalt are poorly exposed in the broader part of the gorge 3 miles upstream from Upper Mesa Falls, where Sheep Falls cuts through an exposed thickness of 25 feet Of a single massive flow of olivine basalt. Presumably the rhyolite—basalt sec- tion of the middle canyon walls at Upper Mesa Falls and the two mixed sections are equivalent. Whether or not any of the rhyolite ash flows exposed in the gorge of Henrys Fork are correlative with ash GEOLOGY AND PETROGENESIS, ISLAND PARK CALDERA, EASTERN IDAHO 013 flows in the rhyolite shield is not known. All the ex- posed flows may be younger than the collapse, having been confined within the caldera by the bounding scarp. Alternatively, the lower part of the rhyolite within the caldera might have been part of the ash-flow sheets that formed the shield and have been dropped when the caldera collapsed. BASALT FLOWS The basaltic, upper part of the caldera fill is well exposed in cuts along the old route of US. Highway 20 and 191 where the road descends from the basalt plain of the fill into the gorge of the Warm River at the south edge of the caldera. All the rock here is simi- lar open-textured light-gray nonporphyritic olivine basalt; it occurs as seven flows, each 4—20 feet thick, totalling 75 feet in thickness. All the flows are of pahoehoe-type lava and have undulating tops, ropy upper surfaces, massive interiors, vesicular upper zones, and less vesicular bottom zones. None of the vesicles are filled. Fresh rock of each flow lies directly on fresh rock of the flow beneath with no trace of erosion, weathering, or deposition between. The upper vesicu- lar zone of one flow has pipe vesicles that are over— turned toward the southeast. Another flow, which is discontinuous in the roadcuts, consists of pods less than 4: feet thick outlined by vesicular zones. The floor of the caldera is a basalt plain (figs. 4, 5, 6) composed of similar pahoehoe flows of open-textured olivine basalt. Most of this floor is uneroded, and only the top foot or two of the upper vesicular zone of the uppermost flow is normally exposed except in roadcuts. Surfaces of the flows are undulating, and the plain defined by the flows undulates correspondingly. Basins in the flow tops are typically filled by loess, whereas domes in many places have smooth, bare-rock pahoehoe surfaces. Local relief of the undulations is generally between 2 and 10 feet, but amplitudes of 20 feet are common, and amplitudes of 50 feet were seen in one area. That this relief is constructional is obvious in all cuts, for the ridges are coincident with pressure domes and the hollows with down-buckles. Surface slopes are as steep as 10°, and no systematic orientation to the undulations was recognized. The larger undula— tions are spaced 500—1,000 feet apart. Squeeze-ups— steeper compressional ridges broken by axial tension cracks and grabens—are rare. Several irregular low scarps in the caldera floor mark abrupt changes in level of 10—20 feet each and have slopes of 10°~15°. These scarps are presumed to be flow fronts, although none is well enough exposed to allow confirmation. No agglomerate or aa lava flows were recognized within the caldera. Swampy Swan Lake, which is west of the center of 7 61—37 4 O—65-———3 the caldera, apparently fills a small collapsed area that is enclosed by a scarp 6 feet high in the surrounding basalt flow. Basaltic cinders were seen within the caldera at only two localities, marked on plate 1 as vents in the west- central part of the floor. The cinders northwest of Lit- tle Butte are yellowish brown and less than 3 inches long. They form a low bill that presumably was a cinder cone. The cinders northwest of Lookout Butte are red and similarly form a low hill. At least four broad low hills within the caldera are composed of olivine basalt: Ripley, Eccles, and Hatch- ery Buttes, and the unnamed butte between Osborne Butte and Henrys Fork. Each is an evenly sloping ba- salt hill that rises above the basalt plain. Eccles and Ripley Buttes are 200L300 feet high and 2 miles across. Hatchery Butte is itself a little smaller but rises from a broad, very gentle dome 5 miles in diameter in the cen- ter of the caldera. Hatchery Butte has slopes typically of about 5°. Eccles and Ripley Buttes are steeper and have extensive 10° and local 15° slopes. All the buttes are densely forested, and no good outcrops were found. All blocks seen on the surfaces are of pahoehoe basalt like that of the basalt plain. No cinders or scoria were seen, and no summit craters were recognized, although the irregular crest of Hatchery Butte is termed a “cra- ter” on the topographic map (pl. 1). These buttes are probably basaltic vent shields, but no convincing evi- dence was recognized. Another possible interpretation is that they are domes uplifted by rhyolite magma which did not break through to the surface. Little Butte, which is near the west edge of the caldera floor, is basaltic according to Stearns, Bryan, and Cran- dall (1939, pl. 3). They also (1939, p. 35) described Olivine Butte, which is near the northwest edge of the caldera and which was not visited during the present reconnaissance, as a tuff cone of olivine basalt erupted through water—saturated alluvium. The cone has a summit crater, and a spatter cone on its flank is regarded as the vent for surrounding flows of basalt. Basalt erupted near the caldera ring fault in the west formed a cinder cone at a vent and a flow sloping to— ward the caldera. Another possible vent is in the raised area 2 miles west of Lookout Butte, near the ring fault in the southwest. Along the east side of the caldera floor, the basalt plain slopes eastward, and its flows pre- sumably came from vents near the center of the caldera. Basalt flowed south through the Henrys Fork-Warm River gap at the south edge of the caldera and is now preserved in benches above the present stream levels. RHYOLITE DOMES Several extrusive rhyolite domes are scattered on the caldera floor. The largest of these are the two domes C14 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY (beyond Silver Lake in fig. 4) adjacent to the ring fault near Bishop Mountain and described earlier. Next largest is Osborne Butte, in the center of the caldera. This butte is a rounded hill 300 feet high composed of phenocryst-rich rhyolite which is mostly lithoidal but which locally contains abundant granules of dark obsidian. Large rounded outcrops of much- weathered rock are common on the south and southeast slopes of the butte down to the level of the basalt plain. Basalt surrounds the hill, but the slopes are entirely of rhyolite. There is no evidence for upbowing of the flanking basalt. The contact between rhyolite and basalt is not exposed, and whether the basalt laps onto an older rhyolite dome or the rhyolite was extruded on a basalt surface was not determined. Four small domes of similar rhyolite rise above the basalt plain near the southwest side of Osborne Butte. Lookout Butte, near the southwestern margin of the caldera floor (fig. 5), is also formed of granular-weathering por- phyritic rhyolite, which crops out in large rounded masses. The butte 2 miles south-southeast of Hatchery Butte consists of rhyolite according to Stearns, Bryan, and Crandall (1939, pl. 3) ; I did not visit it. Soft, pinkish—gray rhyolite tufi' having a gently roll- ing surface forms part of the caldera floor between Lookout Butte and Swan Lake southwest of the caldera center. Unlike most of the rhyolite of the caldera re- gion, this is almost devoid of phenocrysts. RHYOLITE OF THE YELLOWSTONE PLATEAU Rhyolite ash flows and lava flows erupted within the Yellowstone Plateau east of the Island Park caldera form the Yellowstone Plateau and extend into the area shown on plate 1 from the north, east, and south. Three _ subdivisions of these rocks are indicated on the geologic map (pl. 1) : Precaldera ash flows, a postcaldera ash flow, and the lava flows of the Madison Plateau. PRECALDERA RHYOLI'I‘E ASH FLOWS Old rhyolite welded tufi' underlies that part of the area shown on plate 1 southeast of the caldera complex. Extensive exposures were examined during this recon- naissance only in two gorges where crossed by Idaho State Highway 32: the gorges of the Falls River, 4 miles south-southeast of Ashton, and those of the North Fork Teton River, 4 miles south of the southeast corner of the area shown on plate 1. Continuity of the tufi' throughout the area away from the highway traversed is, however, obvious on aerial photographs. The rhyolite exposed low in the gorges of the rivers mentioned and between the rivers in uncommon low outcrops is light-gray to light-purplish-gray welded tufl' containing abundant phenocrysts of sanidine. Both gorges expose probably only the one ash flow. The ex- posed thickness of the flow at the North Fork Teton River is 150 feet. The deep exposures are of densely welded tufl", whereas near-surface exposures high in the gorge walls and on the interfluves are of firmly welded but less dense material. At the Falls River by Idaho State Highway 32, the welded tufi' is cut by three steep north-trending rhyolite dikes, 2, 5, and 50 feet thick, respectively. The interfluve surface—the moderately eroded and variably mantled initial surface of the ash flow—rises regularly eastward to an altitude of about 6,000 feet south of the Falls River and to an altitude of about 5,800 feet north of the river. The unit was traversed only west of these limits. At about the altitudes noted, the slope steepens eastward, then flattens again to a slope like that to the west. The steepening might mark the much-eroded sloping front of an ash flow about 150 feet thick lying upon the flow examined. The upper( ?) ash flow may form the surface of all of the eastern part of the unit within the mapped area. Canyons cut in the eastern part of the unit are as much as 500 feet deep, and layering of the rhyolite, which is visible on aerial photo- graphs, may indicate the presence of separate ash 'flows each 50—200 feet thick. Iddings (1899a) described sev- eral of the eastern ash flows as having lithoidal interiors and glassy, locally agglomeratic bases. The ash flows slope upward to the northeast and dis- appear beneath the young rhyolite lava flows of the Madison and Pitchstone Plateaus of the Yellowstone Plateau. Presumably the ash flows were erupted from vents within the central part of the Yellowstone Plateau and are part of the ash flows that now form the outer part of the Yellowstone Plateau. The ash flows were downfaulted in the interior and buried by younger lava flows (Boyd, 1961). Low, elliptical Rising Butte was not Visited, but its shape suggests that it is a rhyolite lava flow. Basalt flows lie on the tuff in several places but are not separated on plate 1. The basalt also is overlain by the rhyolite lava flows of the Madison and Pitchstone Plateaus (Iddings, 1899a, b; Boyd, 1961). The ash flows are much more dissected than are those of the rhyolite shield of the Island Park caldera and are presumably older; but the contact between them was not seen, and possibly some sheets of the tufl' from the Yel- lowstone Plateau overlie the rhyolite of the caldera shield. The ash flows, as seen in aerial photographs, bend abruptly upward at the northwest base of the Teton Range and form the dip slope of part of the range. The tulf has shared most of the deformation that pro- duced the present range. The hinge line trends north- northeastward from the extreme southeast corner of the mapped area. Within the range the ash flows lap “"1 v '7 | ' IV. ‘I GEOLOGY AND PETROGENESIS, ISLAND PARK CALDERA, EASTERN IDAHO 015 against a rugged erosion surface formed on pre- Pliocene rocks. POSTCALDE‘RA ASH FLOW A rhyolite welded tufl', which probably consists of only one ash flow and which erupted northeast of the Is- land Park caldera, covers the north-central part of the area shown on plate 1. This ash flow, which was erupted after collapse of the caldera, extends uninterrupted southward across the trend of the caldera scarp. The surface of the ash flow rises to the northeast beyond the caldera, and thus a source outside the caldera in that direction is indicated. 'The ash flow is overlain by one of the rhyolite ash flows‘ of the eastern rim, which is overlain in turn by at least two rhyolite lava flows that erupted along the eastern rim of the caldera. The tufi' sheet has been much gullied, and its inter- fluve surfaces have been eroded to a rolling plain. In the walls of the gorge of Henrys Fork from Macks Inn to Island Park Reservoir, the tuff is exposed as a single ash flow. About 200 feet of densely welded light- gray tufl' is exposed at Macks Inn, where flow structures dip irregularly as much as 10° in any direction. Three miles west—southwest of Macks Inn, cliffs 100 feet high expose only the same ash flow. LAVA FLOWS 0F MADISON PLATEA'U Gigantic lava flows of viscous rhyolite that erupted from a fissure zone in the western part of the Yellow- stone Plateau partly cover the easternmost elements of the Island Park caldera and form the northeastern part of the area shown on plate 1. Individual flows have average areas of 150 square miles, thicknesses of many hundreds of feet, and volumes of 5—10 cubic miles. The flows form the Madison and Pitchstone Plateaus, which combined form an area 30 miles long in a north-north— westward direction and 15—20 miles wide. The crestal fissure zone from which all the flows were erupted trends north—northwestward parallel to the long direction of the lava pile and crosses the northeast corner of the area shown on plate 1. Flow fronts are steep, commonly sloping 20° or more for their first few hundred feet of rise. Slopes are more gentle above this, but a mile behind their fronts the flows are 400—1,000 feet higher. The flows rise abruptly above the surrounding area along their composite front. Within the area shown on plate 1, the flows lie on the rhyolites erupted along the eastern rim of the Island Park caldera. Features of the rhyolite lava flows within Yellowstone National Park were described by Hamilton (1960a, 1964). Four lava flows of the Madison Plateau are dis- tinguished on plate 1. The flows were diverted both northward and southward around the plateau formed by the postcaldera rhyolite—lava flows of the east margin of the caldera. They are therefore in general younger than any of the flows of the caldera group, although the small flow remnant east of Buffalo Lake may be older. The flow that forms Black Mountain is one of the caldera group, but it probably is younger than the older flows of the Madison Plateau. The flows which form most of the surface of the Madi- son and Pitchstone Plateaus are of very late Pleistocene age. The flow underlying the northeast corner of the area shown on plate 1 lies on the large terminal moraines of Bull Lake Glaciation in the southern part of the West Yellowstone basin but was cut by ice of Pinedale Glacia- tion (latest Pleistocene) in Madison Canyon (Richmond and Hamilton, 1960). The flow capping the Madison Plateau, which extends almost to the east edge of the southern part of the area shown on plate 1, is similarly of post—Bull Lake, pre-Pinedale age and lies upon a glacially grooved pre-Bull Lake flow (Hamilton, 1960a). Small domes and flows at the crest of the Madi— son Plateau, east of the area shown on plate 1, are of Recent age (Hamilton, 1960a). From oldest to youngest, the four lava bodies that flowed into the map area are designated the northwest- ern, north-central, northeastern, and eastern flows. (These are numbered 1, 2, 3, and 4, respectively, on the geologic map, pl. 1.) The eastern flow largely lacks forest cover and has very well preserved flow ridges. The northeastern flow is older than the eastern flow and also is little eroded, has little soil cover, and has well- preserved flow ridges, but it is completed forested. The north-central flow is still older; it is more eroded and has less rock exposed. The northwestern flow is still more eroded, has relatively few outcrops, and has the thickest cover of loess and soil. The small flow remnant east of Buffalo Lake is interpreted to be part of this northwestern flow. None of these flows, however, show any glacial grooving like that of the pre-Bull Lake flow of the Pitchstone Plateau, and no glacial erratics were recognized on the flow surfaces. Probably they are all younger than the Bull Lake Glaciation, although the Bull Lake ice may instead have been dammed by the east side of the rhyolite flows and been unable to cross this part of the Yellowstone Plateau. Each flow consists of an upper zone of obsidian ag- glomerate about 100 feet thick that grades downward into flow-contorted rhyolite that is several hundred feet thick. Beneath this is massive rhyolite, which rests on an obsidian agglomerate base. The upper agglomeratic zone consists mainly of blocks of black obsidian in a matrix of sandlike un- consolidated glass shards. Exposures (fig. 8) are par- ticularly good in the north—central flow along the road at the north crest of Black Canyon. Obsidian frag- ments range in size from granules to blocks more than SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY FIGURE 8.——0bsidian agglomerate at top of rhyolite lava flow. Blocks of black obsidian lie in an unconsolidated matrix of shards, strands, and granules of clear glass. Shards of the dark matrix at top are coated by ferric oxide dust. Roadcut, 20 feet high, on top of flow at north side of Back Canyon, 2 miles north of area of plate 1. 50 feet in length. The obsidian is minutely fractured and crumbles readily to granules. Many blocks of un- weathered obsidian are so granulated that they can be dug into with a shovel. All the obsidian contains phenocrysts of clear sanidine, and much obsidian is spherulitic. Flow structures are restricted to individ- ual blocks and are oriented chaotically. The rock also contains blocks of pumice, scoriaceous glass, brown ob— sidian, and light-gray lithoidal rhyolite. The matrix is made up of shards, short fibers, and sandlike grains of clear glass that are evenly sized (0.02—02 mm), unsquashed, and unconsolidated. Larger grains are ropy and pumiceous. Most grains are colorless, but some are very light gray or light yel- lowish gray, and small granules of black obsidian are mixed throughout. No welded tufl' was seen. , On the forested surfaces of the flows, abundant ob- sidian blocks project through the soil. The unconsoli- dated matrix is exposed only Where there has been rapid erosion. The older the flow, the fewer the exposed blocks. ‘ Scattered pipelike masses of steeply dipping gray GEOLOGY AND PETROGENESIS, ISLAND PARK C‘ALDERA, EASTERN IDAHO ' fl . Cl7 pumice probably represent upward surges of lava that occurred after differential movement within the ag- glomerate was nearly complete. The agglomerate has been altered locally by fumarolic activity that pre- sumably accompanied cooling of the interior. Siliceous sinter is’ widespread as thin coatings on obsidian blocks. Rarely, hematite and limonite coat the shards of the matrix. Sulfur is abundant in sinter in rare places. The rhyolite flows have been eroded little within the area shown on plate 1, and only their upper, obsidian- agglomerate zones can be seen. The interior of the large flow in the northeast corner of the mapped area is, however, exposed farther northeast along the canyons of the Madison and Firehole Rivers in Yellow- stone National Park (Hamilton, 1964). The flow- contorted rhyolite beneath the agglomerate is well exposed along the Firehole River near Madison J unc- tion and also in clifi's along the south side of Madison Canyon east of Mount Haynes. The underlying mas- sive rhyolite is well exposed also east of Mount Haynes, where the thickness of the flow is 1,000 feet. Boyd (1961) reported a basal obsidian agglomerate at the bottom of a similar flow in Bechler Canyon in the south- west corner of Yellowstone Park. flow domes wnth craters BASALT FLOWS SOUTHEAST OF CALDERA Flows of olivine basalt lie on the precaldera rhyolite ‘ ash flows in several parts of the mapped area southeast of the caldera. Flows having a maximum aggregate thickness of at least 100 feet form a discontinuous veneer on the rhyolite tufl' in the southeastern part of the area and are overlain in turn by the rhyo- lite lava flows in the Madison and Pitchstone Plateaus . (Iddings, 1899a, b; Boyd, 1961). Thin flows of olivine basalt underlie the postcaldera rhyolite ash flow at the eastern rim south of Moose Creek Butte along the middle reaches of North Fork Fish Creek. Inspection of aerial photographs shows that basalt overlies the precaldera rhyolite tufi' north of Robinson Creek. BASALT FLOWS OF SNAKE _RIVER PLAIN Flows of olivine basalt, which include the youngest volcanic rocks within the area shown on plate 1, lap» onto the precaldera rhyolite shield from the west. These basalt flows were erupted from numerous vents that were mostly within a broad west-southwest-trend- ing zone west of the rhyolite dome and flow forming Bishop Mountain (pl. 1 and fig. 9; see also Steams, FIGURE 9.—~Vent zone in basalt of Snake River Plain. Craters and cones of the vent zone, which trends southwestward from west of Bishop Mountain, are of latest Pleistocene or Recent age. Winter view northwestward from above the west edge of the area shown on plate 1, 5 miles southwest of Crystal Butte. 018 SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY Bryan, and Crandall, 1939, pl. 3). The basalt shields— with or without central craters—and cones of the broad vent zone are virtually uneroded (fig. 9). Lava flowed 15 miles southward from the vent zone. The flows are of pahoehoe-type lava near the vents but of aa-type lava near the margins. Aa flows exposed southwest of the caldera are in general successively younger north- ward, in the direction of the main vent region. The sur- face of the youngest flows consists largely of bare jagged rock, whereas the surface of older flows is overlain by considerable loess and, in the southwestern part of the mapped area, by sand dunes. Among the older flows exposed is one that forms the surface north of Bishop Mountain, at the west end of Island Park Reservoir. This basalt flow is similar in preservation to the flows inside the caldera. Broad low outcrops of its original pahoehoe surface coincide with the crests of primary undulations, but the primary ba- sins are largely filled by loess and rubble. SURFICIAL DEPOSITS Unconsolidated alluvial, eolian, and glacial deposits completely cover parts of the area and discontinuously veneer larger parts. The following observations are relevant to interpretation of the age of the rocks and structures of the caldera. Glacial deposits are referred here to the Pinedale (youngest), Bull Lake, and pre-Bull Lake glaciations. The names Pinedale and Bull Lake were introduced by Blackwelder (1915) for the moraines in the Wind River Mountains of Wyoming. Although a confusing variety of other names have been used since for correla- tive materials, Blackwelder’s names have been used in- creasingly in the Rocky Mountain region. Three main advances of ice during Pinedale time and two during Bull Lake time have been generally recognized (for example, Richmond, 1960). In the Yellowstone region the Bull Lake ice was the more extensive and left larger moraines than did Pinedale ice (for example, Black- welder, 1915; Richmond and Hamilton, 1960). Pine- dale moraines are little weathered and have many kettle lakes. Bull Lake moraines are more subdued and gen- erally lack lakes. Pre-Bull Lake till normally lacks obvious morainal form. The use of provincial names such as Pinedale and Bull Lake is unfortunate but is apparently still neces- sary. Although correlation of Pinedale with late Wis- consin of the Great Lakes region, of Bull Lake with early Wisconsin, and of pre-Bull Lake with pre- Wisconsin (Illinoian and older) has long been inferred (Blackwelder, 1915; Richmond, 1962), current disputes regarding status and correlation of early Wisconsin, Illinoian, and intermediate tills even in the Great Lakes area render such correlations uncertain. Pinedale cor- relates with some part of the Wisconsin, but any other correlations are indefinite. Following the conventional use in the Rocky Moun- tain region, Pinedale and Bull Lake are here referred to as within the late Pleistocene; and pre—Bull Lake, as middle Pleistocene and older. Middle Pleistocene thus defined may include much later time than middle Pleis- tocene defined on the basis of marine deposits and fossils. Post-Pinedale time is referred to as Recent. Till older than Bull Lake glaciation and the cutting of the gorges was recognized by Blackwelder (1915) to be widespread on the rhyolite ash flows of the Yel— lowstone Plateau in and south of the southeastern part of the area shown on plate 1. Thick loess overlies this till. Stearns, Bryan, and Crandall (1939) recognized but did not describe similar old till near Warm River community south of the Island Park caldera. This till is well exposed in cuts along the old route of U.S. High- way 20 and 191 southwest of the community; in these places the till contains blocks as much as 5 feet across in a silty matrix. A thick brown soil has developed on this till, and caliche impregnates the upper part of the till. The till overlies the rhyolite ash flows of the precaldera shield. Ice of probable Bull Lake age scoured an old rhyolite lava flow on the Pitchstone Plateau in Yellowstone Na- tional Park, but the scoured surface is overlain by a post- Bull Lake flow which was in turn eroded by ice probably of Pinedale age (Hamilton, 1960a). Most of the rhyo- lite lava flows of the Madison Plateau within the map- ped area are considered to be of post-Bull Lake age, although such an age has been demonstrated only for the eastern and northeastern flows (Hamilton, 1960a; Richmond and Hamilton, 1960). A possible till is exposed in highway cuts on the ba- salt floor of the northern part of the caldera 3.2—3.4 miles south of the Buffalo River. The till( ?) lies on a 40-foot riser that separates two levels of the basalt plain. The deposit consists of angular to subangular blocks of basalt, mostly less than 1 foot across, and sparse cobbles of quartzite, in an unsorted unbedded matrix of silt and fine sand. No soil has developed on the deposit, and very little soil lies on the overlying 1—3 feet of loess although near-surface loess is oxidized. The deposit is exposed through a vertical range of 40 feet, but no more than 15 feet is visible in any section. A low ridge trending westward 100 yards south of this deposit may be a moraine. The ridge consists of a 5-foot-thick deposit of basalt blocks in a silt matrix, overlain by 2 feet of loess. Another low ridge 0.2 mile farther south is formed of gravel composed of well- rounded cobbles less than 4 inches long of basalt and subordinate rhyolite and quartzite in a poorly sorted silty matrix; it may be an esker or a kame terrace. N0 an I} .r. GEOLOGY AND PETROGENESIS, ISLAND PARK CALDERA, EASTERN IDAHO 019 3‘ other materials of possible glacial origin were recognized 5 within the caldera. The lack of soil on these possible glacial deposits sug- ‘~ gests that they may be of Bull Lake or younger age. As Pinedale glaciers extended little beyond the mouths of the major deep canyons draining the west side of the Yellowstone Plateau, (Richmond and Hamilton, 1960), these deposits must be of pre-Pinedale age. They are about as far from the Yellowstone Plateau as are the known Bull Lake tills of the southeastern part of the mapped area, and correlation with Bull Lake is sug- gested. Ice of Pinedale age scoured the youngest large rhyo— lite lava flows of the Madison Plateau (including the eastern and northeastern flows of the mapped area) and Pitchstone Plateau, but small flows and domes on the Madison Plateau east of the mapped area are post- Pinedale (Richmond and Hamilton, 1960; Hamilton, 1960a). River gravels and finer alluvium are widespread along river flats, notably those west of Last Chance within the caldera and near Ashton, and on terraces above the rivers. The highest gravels seen lie 400 feet above pres- ent stream level on the divide between the Warm River and Henrys Fork. Loess covers large areas of the river deposits. Broad swampy areas in the northern part of the caldera suggest continuing subsidence. The alluvial deposits east of Macks Inn near the north edge of the caldera and those in the southeastern half of the large alluvial area (pl. 1; fig. 7) that extends from Ripley Butte to Island Park Reservoir consist of obsidian sand. Granules of black obsidian lie in glass sand. This material was derived from the unconsoli— dated obsidian agglomerate upper zones of the nearby rhyolite lava flows. Loess is widespread on all volcanic units of the mapped area except the younger of the basalts west of the caldera. Ten feet or more of loess is common on the rhyolite ash flow southeast of Ashton. Loess sev- eral feet thick characteristically lies on the precaldera rhyolite shield and on the other postcaldera rhyolite ash flows and lava flows of the eastern rim. Loess also lies between two tufis in the rhyolite shield. The loess layer is generally thinner and less continuous on the rocks of the caldera floor and on the younger rhyolite lava flows of the eastern rim and of the Madi- son Plateau. Dunes of loess and sand occur on the older basalt flows near the southwest corner of the mapped area. AGE OF VOLCANIC BOOKS The oldest unit of the mapped area is probably the rhyolite-ash-flow sequence of the Yellowstone Plateau that is exposed southeast of the caldera. This sequence is not younger than middle Pleistocene. Love (1956) found the rhyolite tufl's in the Teton region to be of Pleistocene and early, middle, and late Pliocene ages; but the tuf‘f within the mapped area apparently has not been dated except as older than the pre-Bull Lake glaciation. The precaldera rhyolite shield also is older than the pre-Bull Lake glaciation and, therefore, is not younger than middle Pleistocene. It is, however, so little eroded that substantially greater age is unlikely, and accord- ingly it is tentatively assigned a middle Pleistocene age. The floor of the caldera lacks pre-Bull Lake till; hence its basalt is younger than that glaciation. A possible Bull Lake till lies on basalt in one part of the caldera; if this deposit has been identified correctly, then that ba- salt is of late middle or early late Pleistocene age. Ba— salt in the western and southern parts of the basalt plain has broad areas of outcrop and, in places, unfilled squeezeups; it may be of late Pleistocene age. Probably all the rhyolite lava flows of the Madison Plateau within the mapped area and the youngest post- caldera rhyolite lava flow of the eastern caldera rim were erupted in the interval between the Bull Lake and Pinedale Glaciations and are thus of late Pleistocene age. The other elements of the caldera are bracketed be- tween the middle Pleistocene age of the precaldera shield and the late Pleistocene age of the rhyolite flows. Collapse of the caldera probably occurred in middle or late Pleistocene time. The basalt flows of the Snake River Plain west of the caldera shield include units that almost totally lack loess cover or any modification of jagged flow surfaces or vent structures (fig. 9). These young flows are prob- ably of Recent age. Other flows, their surfaces now somewhat subdued and their depressions filled by loess, are considered to be of late Pleistocene age. The basalt of the caldera floor is younger than the rhyolite of the caldera shield, but it is probably older than the postcaldera rhyolite ash flow (fig. 6) and the lava flows of the eastern caldera rim. PETROLOGY The rhyolites of the Island Park caldera and the ad— jacent Yellowstone Plateau vary widely in appearance but little in composition. All are highly silicic rocks containing phenocrysts of sanidine, quartz, and, com- monly, oligoclase, but containing almost no mafic min— erals. Precollapse and postcollapse rocks, and those from Island Park and Yellowstone Plateau sources, are indistinguishable compositionally. Basalt of the caldera fill is holocrystalline olivine- augite basalt that also varies little in composition. Ba- salt flows outside the caldera are of different olivine— a C20 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY augite basalt. Except for a single mass of latite, the rocks of the caldera suite are completely bimodal—— silicic rhyolite and olivine basalt. PETROGRAPHY RHYOLITE The rhyolite occurs as tuffs, mostly welded, and as flows and domes. All forms contain phenocrysts or crystal fragments of clear sanidine and subordinate high—quartz, and most also contain crystals of oligo— clase. Mafic minerals total much less than 1 percent of any specimen; green augite and variably altered fayalite are common, and biotite and muscovite are rare. The most abundant rocks are light-gray, light- purplish-gray, and pale—red welded tufl's. They are massive to streaky rocks of lithoidal aspect despite their almost wholly vitric groundmass. Although the rocks vary in hardness from those which ring when struck with a hammer to those which thud and crumble, all those examined in thin section show flattened—shard structure (fig. 10A) such as that described by Boyd (1961) and by Ross and Smith (1961) as typical of welded rhyolite ash flows. Welding in the denser rocks has variably obliterated shard outlines (fig. 103). Shards are now lenses a few tenths of a millimeter long, oriented parallel in crystal-poor rocks but irregularly in crystal—rich ones (fig. 100). Lenses of squashed pumice or, less commonly, of devitrified rhyolite streak many rocks. Groundmasses of most are completely or largely glassy. Devitrification is generally restricted to tiny lenses of collapsed pumice or other enclosed rhyo- lite clasts. Rhyolite from the domes examined around the western and southern caldera rim is mostly very light gray, chalky—looking rock that contains a few percent of small (1 mm) phenocrysts of sanidine, high-quartz, and Oligoclase. Groundmasses are composed of micro- crystalline quartz and alkali feldspar in hazy inter- growths; one specimen has irregular patches of microgranophyre. Biotite, which is lacking in the welded tufts, is a rare component of dome rocks. De- vitrification spherulites are abundant in once-glassy rocks on the crest of the Bishop Mountain dome and flow (fig. 10D). The rhyolite of a small dome, OS— borne Butte, in the center of the caldera contains more and larger phenocrysts than does the rhyolite of the rim domes, and it is similar in this respect to much of the welded tuff of the area. Variably devitrified and spherulitic black obsidian forms abundant blocks in the agglomerates at the top of the rhyolite lava flows. The obsidian where undevitri- fied is granular and crumbles readily. A few to 10 per- cent of the rock is phenocrysts. The most dissected flows contain the smallest amount of glassy obsidian and the greatest amount of gray devitrified rhyolite. Lith- oidal and glassy rhyolite blocks in the upper part of the flows vary greatly in texture and appearance. Flow structures defined by microlites pass without interrup- tion through spherulites, which indicates that the spherulites are products of devitrification. The mineralogy of the various textural types of rhyo- lite is similar. Sanidine is in clear, generally euhedral grains that are 1—5 mm long, are commonly embayed, and have a —2V of 10 °—20°. These optic angles indicate a composition of about Or62_70Ab23_36An2 according to Troger’s (1956, p. 96) tables. Quartz forms unstrained bipyramids as much as 2 mm long. Tridymite occurs as feathery sheaves. Oligoclase shows albite twinning but generally shows no zoning. Clinopyroxene occurs as pale-green equant granules and as sparse microlitic needles; it has large extinction angles and large optic angles. Fayalite ( —2V=50°~ 60°) is in euhedral microprisms and ranges from fresh to completely altered. The rare biotite is light brown and generally is altered. No amphibole was seen. A few magnetite granules are present in most thin sections, and some granules have hematite rims. How- ard A. Powers (oral commun., 1961) crushed a large sample (table 1, no. 5) and found chevkinite (a Ca-Fe- Ti-rare earth silicate, pleochroic from reddish brown to opaque) to be abundant in the sparse heavy concentrate derived from it. BASAL’I‘ The basalt flows of the caldera vary little in mineral composition and texture. Most of the basalt is open textured, diabasic, and nearly holocrystalline (fig. 11 A). Its dominant mineral is calcic labradorite in thin laths about 0.5 mm long. These are enclosed in variably ophitic plates of pale-olive or pale-green augite that has a small optic angle (+2V=30°—40°), which indicates a composition that probably is high in magnesium and low in calcium. Prisms and granules of forsteritic oli— vine (2V=90°) are partially altered in some specimens. No pigeonite or orthopyroxene was noted. Ilmenite occurs as plates, and magnetite as granules. Plagio- clase crystals project into the abundant tiny intergranu- lar cavities and appear white, giving the rocks their light color. A basalt specimen from Hatchery Butte is markedly different in texture and contains plagioclase laths and aggregates of olivine granules in a fine—grained ground- mass of plagioclase laths, clinopyroxene granules, opaque minerals, and minor cryptocrystalline matter. The one basalt specimen studied from the Snake River Plain outside the caldera (table 1, No. 1) is olivine- V V GEOLOGY AND PETROGENESIS, ISLAND PARK CALDERA, EASTERN IDAHO 021 4 2% mm FIGURE 10.-—Photomicrographs of rhyolite. A, Welded rhyolite tuft. Squashed shards in rhyolite ash flow of caldera floor, from highway cut 3 miles east-southeast of Little Butte. B, Welded rhyolite tuft. Refusion has obliterated the outlines of many of the squashed shards. The large crystal fragment above the scale is sanidine. Postcollapse ash flow of the eastern caldera rim, by forest road at southeast base of Moose Creek Butte. 0, Crystal-rich rhyolite welded. tuft. The shards are deformed around the crystals of sanidine, quartz, and oligoclase. Many squashed fragments of micro crystalline rhyolite and partially divitrified pumice. Analyzed specimen 5 (table 1), rhyolite of caldera shield, from highway cut 6 miles north-northeast of Ashton. D, Devitrification spherulite in Lobsidian. The large crystal of sanidine (clear, right and bottom) has been partially resorbed and converted to microcrystalline material. Analyzed specimen 6 (table 1), obsidian block on surface of rhyolite dome, Bishop Mountain. augite basalt that is markedly different from the caldera basalt. The pyroxene of this basalt is brown, has a large optic angle (+2V~65°), and is probably more calcic than that of the basalt of the caldera. The plagio- clase is calcic andesine (Ana); so the rock might be designated basaltic andesite. The only common opaque mineral is ilmenite, which occurs as thin plates. The texture of this rock, as of the caldera basalt, is diabasic and almost holocrystalline (fig. 113). The one specimen of latite collected from Antelope Flat on the caldera rim is described in table 1 (No. 4) . CHEMISTRY Chemical analyses for both major and minor elements were made of 11 specimens of volcanic rocks from the Island Park caldera and Vicinity (table 1). Included were two analyses of basalt from the caldera fill, one of basalt from outside the caldera, one of latite, and seven of rhyolite from ash flows and domes. The volcanic assemblage is remarkably bimodal and consists of olivine basalt on the one hand and highly silicic rhyolite on the other. The one specimen of latite analyzed is from the only area of intermediate rocks C22 SHORTER CONTRIBUTIONS . ~ , \ . we.” r , , ‘V: W 3,, '1 _ . m . w ' \ sv- » ’ W¢$‘Z‘ T0 GENERAL GEOLOGY FIGURE 11.—Photomicrographs of basalt. A, Diabasic olivine basalt of caldera fill. Labradorite laths are enclosed in sub- calcic augite. Three equant granules of olivine are in upper left. Analyzed specimen 3 (table 1), from gorge of Henrys Fork at Upper Mesa Falls. B, Diabasic olivine basalt of Snake River Plain. A phenocryst of labradorite with attached olivine granules lies in a nearly holocrystalline groundmass of plagioclase laths and augite plates. Analyzed specimn 1 (table 1) from a flow near Blue Creek, southwest of caldera. seen in the field; all other rocks noted are either basalt typical of most olivine basalts elsewhere. (For com— 1 or rhyolite. parisons, see Green and Poldervaart, 1955.) Sodium, BASALT potassium, titanium, and phosphorus, however, are ‘ The basalt of the caldera (table 1, Nos. 2, 3) contains markedly less abundant in this TOCk than in typical *. silica, total iron, magnesium, and calcium in amounts olivine basalt. These four elements are less abundant NOTES TO TABLE 1 l. Olivine-augite basalt from Snake River Plain outside caldera. Diabasic texture Collected beside road to Bishop Mountain Lookout near south flank of flow 1 (fig. 118); nearly holocrystalllne; laths 0.5 mm long of plagioclase (A1150) partly north of Antelope Flat. 4 enclosed in plates of brown augite (+2V'»:65°); olivine (2V~90°) granules are 7. Lithoidal rhyolite from Warm River Butte dome. Pinkish-gray rock containing about 7 percent phenocrysts in a massive aphanitic groundmass. Phenocrysts unaltered; ilmenite in plates; cryptocrystalline brown material minor. Vesicular ‘ medium-gray rock from top of an aa flow of Recent age that laps onto the rhyoiite are euhedral crystals 1-4 mm long of sanidine (~2Vfi 20°) and subordinate . shield on the southwest near Blue Creek. oligoclase and quartz. Matrix is a microcrystalline to granophyric mush of ‘ 2. Olivine-augite basalt of caldera fill. Open textured, diabasic, holocrystalline. quartz and alkali feldspar crystals 0.01—0.05 mm in diameter; mush also contains Labradorite (A1105) laths 0.5 mm long are partly enclosed in plates as much as sparse granules of magnetite. There are sparse crystals of much-altered light- I 1.5 mm long of pale-olive augite (+2Vz35“); olivine (2Vz90°) granules as much brown biotite and wavy-extinction muscovite and a few completely altered as 1.5 mm long, altered along margins and cracks; ilmenite plates and magnetite microphenocrysts of an unidentified maflc mineral that has been replaced by granules; no cryptocrystalline material. Medium-light-gray nonvesicular decussate biotite containing granular magnetite and hematite. Numerous fl basalt, having slight purplish cast, from the middle of pahoehoe flow at the top of the section in the gorge of Henrys Fork at Upper Mesa Falls, near junction of highway and road to falls. . Olivine-augite basalt of caldera fill. Open textured, diabasic, holocrystalline (fig. 11A). Labradorite (Arm) laths 0.5 mm long are partly enclosed in plates asmuch as 1mmlong of pale-green augite (2V=30°); 1=mm granules (2V=90°) are unaltered; ihnenite plates and magnetite granules; no cryptocrystalline material. Medium-light-gray nonvesicular basalt, having olive cast, from 10 ft above base of lowest pahoehoe flow in gorge of Henrys Fork at Upper Mesa Falls, south of falls parking area, above rhyoiite welded tufi that forms the falls. elliptical sheaves of tridymite. 8. Welded rhyoiite tufi from caldera fill. Medium-light-gray rock. Crystal rich; contains about 20 percent crystals and fragments 1-5 mm long of clear sanidine and subordinate amounts of oligoclase and high-quartz. Many crystals of green clinopyroxene; some of magnetite. Groundmass is largely glass but is partially devitrified and shows complete squashing and welding of shards. Collected at top of Upper Mesa Falls from thick ash flow in gorge of Henrys Fork. 9. Welded rhyoiite tufi from caldera fill. Light-purplish—gray rock. Contains about 3 percent crystals and crystal fragments of sanidine, subordinate high-quartz, and traces of clinopyroxene and magnetite in a vitric, completely welded tufi . Latite from mafic lava flow on western part of caldera shield. Seriate micro- ‘ phenocrysts 0.1—2 mm long of andesine lie in a very fine grained matrix of flow- matrix. The squashed shards are about 0.2 mm long and are very thin. The alined oligoclase (Anzo) laths 0.05 mm long, clinopyroxene granules 0.01 mm in crystals are oriented parallel to flattening of shards. Collected from ash flow ‘ diameter, tiny anhedra of alkali feldspar, opaque granules, brown glass and in section of interlayered rhyoiite and basalt (above samples 3 and 8 but below 2 brown cryptocrystalline material. Sparse microphenocrysts of olivine (about and 11) in gorge of Henrys Fork above Upper Mesa Falls. F070) and of a green clinoamphibole having a small extinction angle. Dark—gray 10. Welded rhyolite tufi from postcaldera ash flow north of caldera. Light-purplish- rock containing abundant flow-alined vesicles, from Antelope Flat. gray rock that has conspicuous horizontal structure formed by squashed-pumice . Welded rhyoiite tufl from ash flow of the precaldera shield. Pale-red rock contain- lenses. Contains only about 1 percent sanidine crystals about 1 mm long and 1 ing about 10 percent euhedral and fragmental crystals mostly 1—2 mm in diameter; rare high-quartz, clinopyroxene, and altered fayalite(?). Matrix shards are clear sanidine (—2V=10°) is more abundant than are bipyramids of high-quartz; completely squashed and welded and are almost entirely vitric although ‘ minor unzoned oligoclase; sparse magnetite granules with oxidized rims; rare speckled by feldspar microlites. Collected by Henrys Fork from cliffs on upper l altered fayalite (—2~60°) and chevkinite. Matrix is completely welded and part of a thick ash flow that was erupted probably from a vent in the Yellow- { partly refused vitric tufl showing secondary flowage; contains many squashed stone Plateau. fragments of pumice and devitrlfied rhyoiite (fig. 100). Collected in caldera 11. Welded rhyoiite tufi of caldera fill. Pale-red rock streaked by gray lenses of shield from out on US Highway 20 and 191- collapsed pumice typically about 0.5 mm thick and 5—10 mm long. Contains ' Spherulitic rhyoiite from BiShOD Mountain dome. Light-reddish-gray fléw' about 5 percent crystals and fragments, mostly of sanidine and quartz; sparse ( layered rock containing abundant. 1:3 mm spherulites (fig‘ 10D). Contalns pseudomorphs after fayalite(?). Recognizable shards are few and tiny: re- about 5 percent phenocrysts of sanldlne (—2Vz10°) and hlgh-quartz, and rare . , biotite. Pyroxene microlites show flow structure in the groundmass that con- flowage has largely obllterated clastlc texture. Collapsed-pumice lenses show * tinues without interruption through spherules: the spherules are devitriflcation comb-structure devitriflcation, but rest of rock is largely glass. Collected from products. Groundmass consists of devitriflcation microgranules and feathery higheSt 85h flow in section in gorge 0‘ Henrys Fork at Upper M953 Falls, below , 4 aggregates of alkali feldspar and quartz and numerous tiny sheaves of tridymite. capping sequence of basalt flows. -P‘ ..r44 '— GE‘OLOGY AND PETROGENESIS, ISLAND PARK CALDERA, EASTERN IDAHO in most tholeiite than in most olivine basalt; the sodium, titanium, and phosphorus of the Island Park rocks are in amounts characteristic of tholeiite, whereas potas- sium is uncommonly scarce even for tholeiite. The ba— salt in the Island Park caldera thus presents an unusual combination of components. The basalt of the Snake River Plain outside the cal- dera is represented by only one analysis (table 1, No. 1). This basalt difl'ers markedly from that inside the caldera in its content of most major components except silica (fig. 12) and more generally resembles typical olivine basalt. This basalt contains three times as much potas- sium, twice as much titanium and phosphorus, and a little more sodium than does the basalt inside the cal- dera. The basalt outside has a higher content of total iron and lower contents of aluminum, magnesium, and calcium than does the basalt inside. C23 These differences in chemical composition are reflected in the mineralogical contrasts. The basalt inside the caldera has more magnesian pyroxene and more calcic plagioclase than does the basalt outside. The basalt outside the caldera has brown augite that presumably contains considerable titanium, and it has ilmenite as its only conspicuous opaque mineral. Both basalts, how- ever, are properly classed as olivine basalt. Spectrographic analyses are too few for statistical reliability, but they reveal that contents of some minor elements appear to vary in parallel fashion with those of major elements. Reported as more abundant in the basalt outside the caldera are barium (with potas- sium?), cobalt (with ferrous iron?), niobium (With ti- tanium ? ), strontium (affinity for potassium being more significant here than affinity for calcium '9 ) , and yttrium, ytterbium, and zirconium. On the other hand, the vari- TABLE 1,—0hemlcal analyses of volcanic rocks from Island Park caldera and vicinity [Major-oxide analyses made by standard silicate methods by Dorothy F. Powers, Denver, Colo., 1962; fluorine and chlorine analyses by Elaine Munson, Denver, 1962; minor—element analyses made by semiquantitative spectrographic methods by J. C. Hamilton, Denver, 1961, and reported as midpoints of logarithmic-sixth divisions. Location of specimens is marked on the geologic map (pl. 1)] Basalt flows Rhyolite Latite 'I‘ufi of Caldera rim Average Outside Inside caldera Caldera Tufl eru ted during and after of samples caldera rim 001 pse oi caldera 5-11 Flow Dome 1 2 3 4 5 6 7 8 9 10 11 12 ‘1 Major oxides, in weight percent 1 ‘ 47. 35 47. 22 47. 61 61. 33 75. 67 77. 08 77. 22 74. 79 75. 88 75. 63 76. 53 76. 1 14. 55 16. 16 16. 07 16. 34 12. 69 12. 16 11. 90 12. 71 12. 50 12. 54 12. 14 12. 4 1. 99 2. 55 1.22 3. 45 1. 52 . 86 l. 36 1.41 1. 34 1.35 1. 52 1. 3 12.58 9.54 9.85 4.50 .29 .36 .16 .72 .45 .30 .08 3 6. 28 8.46 8. 68 . 47 07 . 08 . 08 . 13 . 08 . 03 03 07 8. 93 11.02 11.28 2. 48 59 . 48 . 53 .84 .57 . 45 51 6 2. 62 2. 44 2. 26 4. 44 3 27 3. 29 3. 24 3. 35 3. 50 3. 59 8 15 3 3 . 83 . 17 .37 4 27 5 32 4. 97 5. 07 5. 25 5. 01 5. 19 5 18 5 1 3. 45 1.68 1.77 72 18 . 15 . l7 . 23 . 17 . 16 18 18 .62 .20 .30 19 02 .01 .01 .02 .03 .01 .01 .02 .22 .20 .20 19 04 .03 .02 .04 05 .02 .02 .03 .32 .17 .30 56 26 .29 .20 22 15 .28 .17 } 3 08 . 10 . 13 56 08 . 06 . 05 19 08 . 18 . l2 . 00 . 00 01 . 00 . 00 01 00 . 00 . 00 00 01 . 01 . 01 03 01 . 05 01 02 01 . 04 01 02 06 . 02 . 03 02 01 . 06 02 02 01 . 01 01 02 Subtotal. 99. 89 99. 94 100. 08 ' 99. 55 100. 03 99. 93 100. 04 99. 95 99. 83 99. 78 Less 0. .03 . 01 . 01 . 02 . . . 01 . 01 . 00 . 01 Total. . 99. 86 99. 93 100. 07 99. 53 100. 03 99. 89 100. 03 99. 94 99. 83 99. 77 ‘ Minor elements. in weight percent C24 SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY Rhyolite tuff Rhyo|ite dome and flow Basalt Latite f 15 A1203 10~————*———~————~——4 MgO |||l||ir| X ,_ O llllllllllllll M ’11 m as FeO Illlllllllllll OXIDE, IN PERCENT BY WEIGHT llll||lllr|l x- llllllllllll z .5“ o Illl X llll 74 N O l||||l llllll l l | l l l | l l I l I : X'X d N0 Ill | J l l l ’ | l .1-.. 45 5O 55 60 65 7O 75 80 SiOZ, IN PERCENT BY WEIGHT 450 0 FIGURE 12.—Silica-variation diagram for rocks of Island Park caldera. Data from table 1. X, basalt from outside the caldera. ations reported in contents of chromium, gallium, and vanadium are opposite to those predictable in terms of common ionic substitutions. Chlorine is equally abundant in all three basalt speci— mens, but fluorine is markedly more abundant in the basalt outside the caldera. The basalt inside the caldera difl'ers most conspicu- ously from that outside by having smaller contents of the more volatile elements, both major and minor, and larger contents of the more refractory elements. Water contents of both rock types, however, are equally low, and neither shows appreciable mineral alteration. LATITE The one specimen of latite analyzed (table 1, NO. 4) has a large content of K20 and NagO (each more than 4 percent by weight) and a small content of silica (61 per- cent). The rock also has a high content of iron, much of it oxidized. Barium (0.5 percent), niobium, zirco- nium (0.1 percent), and the rare earths and allied ele- ments (cerium, lanthanum, neodymium, yttrium, and ytterbium) are all uncommonly abundant in the latite. RHYOLITE The seven specimens of rhyolite analyzed represent the precaldera ash-flow shield (table 1, NO. 5), flOWS and domes on the shield (Nos. 6, 7), ash flows of the caldera fill (Nos. 8, 9, 11), and a postcollapse ash flow erupted in the Yellowstone Plateau (No. 10). These rhyolites vary remarkably little in composition, as is shown by their ranges: Weight percent SiOz ___________ 74.8 —77. 2 A1203 __________ 11. 9 —12. 7 Iron as Fe0_-__ 1.1 — 2.0 MgO __________ .03— .13 CaO ___________ .45— .8 Weight percent Na20 __________ 3.1 — 3.6 K20 ___________ 5. 0 —— 5. 3 T102 ___________ .15— .23 P205 ___________ .01— .03 MnO __________ .02— . 05 Despite the narrowness of this compositional range, the contents of aluminum and total iron in the rocks decrease systematically with increasing silica, and all other major oxides show slight and irregular decreases (fig. 12). This contrast indicates that had the rhyolites crystallized, the least silicic would contain substantially more iron oxide and iron-rich mafic minerals but only slightly more feldspar than would the most silicic rocks. Phenocrysts and crystal fragments range in abun- dance from 1 to 25 percent of the rhyolite. Among the analyzed rocks (table 1), NO. 10 has 1 percent total crystals, No. 8 has 20 percent, and the others each have between 3 and 10 percent. Chemical compositions show no variation reflecting this range: the bulk composition of groundmass and phenocrysts must be nearly the same. The only possible exception is NO. 8, which has slightly more aluminum and calcium (though not sodium) than have most of the other rhyolites; this possibly indicates enrichment in crystals of Oligoclase. Potassium is dominant over sodium by weight but is about equal to it in molecular proportion. The molecu- lar ratio K20 : NazO : CaO in the average rhyolite (ta- ble 1, No. 12) is about 48: 44: 8. The molecular sum of these components is almost exactly equal to that of alu— mina. The two dome and flow rocks have slightly less potassium than do four of the five specimens of welded tufl". Iron in the rhyolite is mostly oxidized; the ratio Fe203 : FeO by weight ranges from 2.4: 1 to 19: 1. The lowest ratio is in the rock of one of the flows (table 1, NO. 6) ; this same rock also has the highest reported con- tents of chlorine and fluorine. These features perhaps indicate that the rock has undergone minimal mixing with atmospheric oxygen and has lost a minimum of magmatic volatiles. The slight variation in water con- rm; ' Q J' l GEOLOGY AND PETROGENESIS, ISLAND PARK CALDERA, EASTERN IDAHO C25 tent among speci ens has no apparent correlation with occurrence, and aEare dry rocks. Of the minor lements, barium is clearly less abun— dant in the flow and dome rocks than it is in the tufl's; and copper, lanthanum, neodymium, phosphorus, stron— tium, and zirconium seen to be less abundant also, al- though the data for them are statistically inconclusive. The more volatile of the rock-forming components— potassium and the elements just listed—seem to be slightly concentrated in the ash flows relative to the domes. . INTERPRETATION OF THE ISLAND PARK CALDERA ASH FLOWS Welded rhyolite ash-flow tufi is a major component of the caldera edifice. Recent literature on origin of such tuff is extensive—among the more significant papers are those by Boyd (1961) and by Ross and Smith (1961)—and no detailed review is given here. The rhyolite ash flows form with smooth upper sur- faces that slope but a few degrees, and at least one of the ash flows ends at a steep front (fig. 6). Such ash flows form from material that flowed as a dense fluid and not from ash carried in a cloud and rained from the sky. The fluid of the ash flows must have had two phases—liquid and gas—for the squashed-shard struc- ture seems inexplicable in other terms. Such a two— phase fluid can be pictured as a pyroclastic flow contain- ing nearly molten glass shards dispersed in gas, as a froth flow containing abundant gas dispersed in nearly molten lava, or as some intermediate between the two. The froth-flow concept seems to explain better the steep front and the smooth, evenly sloping upper surface; but the upper, lowest pressure part of a froth flow might be expected to be like a pyroclastic flow. Very rapid eruption seems required. PHYSICAL HISTORY The Island Park caldera probably formed simul- taneously With the building of the broad rhyolite shield which it displaces. The repeated virtually instantane- ous eruptions of the large rhyolite ash flows which pro- vided most of the substance of the shield must have re— sulted in collapse of the roof of the large magma cham- ber from which the flows came. The collapse presum— ably occurred in increments that were synchronized with eruptions, and the accumulation of loess between two tufl’ units in one part of the rhyolite shield suggests that in— tervals thousands of years long separated some erup— tions. The magma chamber was at least as large in plan as the caldera, 18 by 23 miles, near the surface, but may have narrowed downward. ’ The collapsed part of the caldera was filled first by rhyolite whose eruption was probably synchronous with the collapse and with the building of the rhyolite shield around the caldera. Subsequent eruptions Within the caldera floor were of both basalt and rhyolite, but al- though small domes of rhyolite were erupted late in the period of caldera filling, final eruptions were mostly of basalt. This age sequence in magmas erupted from vents interspersed about the caldera floor indicates that the single large magma chamber probably contained both basalt and rhyolite magmas simultaneously, and that the rhyolite magma probably overlay the basalt magma, for it generally erupted first. Domes and lava flows of rhyolite were erupted near the caldera scarp before final collapse, at least on the western rim. An ash flow and subsequent lava flows were erupted from the eastern rim after collapse and after much of the caldera fill had been erupted. Eruptions of gas—charged rhyolite that produced ash flows were apparently followed in several instances by eruptions of gas—poor rhyolite that formed domes and thick lava flows. The two big rhyolite lava flows and interbedded ash flows of the southwestern part of the shield may represent viscous eruptions following one of the ash-flow-and-collapse episodes. The domes of the western rim were extruded on the ash flows and were themselves broken by the caldera fault. Following this ash-flow-and-dome cycle, the basalt in the caldera floor was erupted. Still later, the eastern rim and all but possibly the youngest part of the caldera fill were buried by the products ofanother rhyolite cycle in which erup- tion of a large ash flow preceded eruption of lava flows. The building of the exposed elements of the caldera has occupied the latter part of Quaternary time—per- haps the last million years. The youngest volcanism within the caldera probably occurred in pre—Pinedale time, but future volcanism is likely to occur. Basalt has been erupted Widely just west of the caledera during Recent time. PETROGENESIS The Island Park caldera is made up of rhyolite and olivine basalt erupted from a single large magma cham- ber. Progressions of eruptions with time suggest that, in general, volatile—rich (ash flow) rhyolite magma lay above volatile-poor (lava flow and dome) rhyolite magma, and that both these rhyolite liquids lay above basalt liquid in the magma chamber. This layering within the magma is interpreted to be the product of liquid fractionation of initially tholeiitic basalt magma into contrasted rhyolite and olivine basalt magmas. This process is suggested to be generally ap- plicable to provinces of bimodal rhyolite-and-olivine- basalt volcanism and is discussed in detail following the presentation of data from other such provinces. C26 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY GEOLOGIC AND PETROLOGIC COMPARISONS YELLOWSTONE PLATEAU' Rhyolite lava flows and ash flows of the Yellowstone Plateau overlap the Island Park caldera from the east. Rhyolites of the caldera and plateau suites are virtually identical chemically, the basalts of the two suites are closely similar, and the occurrences of the two suites are so alike that related origins are required. The plateau is about 50 miles across and is surfaced largely by rhyolite, which fills a broad basin defined by high mountains to the north, east, and south, but spills into the Island Park region to the southwest. The outer part of the plateau, except in the southwest, is formed largely of enormous ash flows of rhyolite erupted during Plio— cene and Pleistocene time (Boyd, 1961; Love, 1956; Hamilton, 1960a, 1964). Collapse of the interior of the plateau—a consequence of the eruption of the ash flows—during Pliocene( ?) and Pleistocene time pro- duced an irregular depression 30 miles across (Boyd, 1961) that has since been flooded by the huge viscous rhyolite lava flows of the Central, Madison, and Pitch- stone Plateau sectors of the Yellowstone Plateau (Boyd, 1961; Hamilton, 1960a, 1964). All rhyolites——ash flows, lava flows, and domes—of the Yellowstone Plateau vary little in mineralogy and less in chemistry. Phenocrysts of sanidine and high- quartz are abundant throughout the rocks, and those of oligoclase are common; mafic minerals, mostly faya- lite and clinopyroxene rather than biotite or hornblende, are sparse (Iddings, 1899a; Boyd, 1961; Hamilton 1964). The uniformity in composition of the rhyolite is shown by the average and range of eight analyses given by Hamilton (1963a) for all types of rhyolite from the northwestern part of the plateau: Weight percent Average Range SiOz __________________________________ 76. 4 75. 7—77. 3 §lgga _________________________________ 12. 2 11. 9—12. 7 e2 3 _________________________________ . 9 . 9— 1. 9 F 0 . 6} as FeO . 13 . 07— 0. 2 4 . 2— 0. 6 3 3 3. 0— 3. 6 4 9 4. 6— 5. 0 6 . 3— 1. 1 . 13 . 1- 0. 2 . 02 . 01— 0. 4 04 . 02— 0. 06 . 08 . 01— 0. 14 . 05 . 01— O. 09 (Eleven other published modern analyses give an al- most identical average but a slightly greater range—— from 73.1 to 78.0 percent SiO2 [Hamilton, 1959]; the wider range may reflect analytical biases rather than true variation in the rocks.) Welded tufl's analyzed con- tain less fluorine and chlorine and more highly oxidized iron than do lava-flow or dome rocks, but the rocks are otherwise indistinguishable compositionally from each other and from the rhyolite of the Island Park caldera. Basalt is greatly subordinate to rhyolite among the exposed rocks of the Yellowstone Plateau occurring mostly as small flows on the rhyolite ash flows of the outer part of the plateau. The basalt varies from dia- basic olivine basalt to basalt andesite (Iddings, 1899b; Wilcox, 1944; Boyd, 1961; Hamilton, 1960a, 1964). The average and range of seven analyses of basalt from Fenner (1938; three analyses of one flow are weighted as a single analysis) and from Hamilton (1963a) are Weight percent Average Range 8102 _________________________________ 49. 3 46. 7—51. 5 A1203 _________________________________ 15. 6 15. 2—15.9 F6203 _________________________________ 2.3 } 10.4—14. 1 FeO __________________________________ 9. 7 as FeO MgO _________________________________ 6.9 6.0— 7.8 02.0 _________________________________ 9. 6 8.5—11. 1 NaéO _________________________________ 2.9 2. 34— 3.3 a __________________________________ . 6 . — 1. 1 H,O __________________________________ .4 .3- 1.2 gig, _________________________________ 2. 0 1.;- 2.: 2 5 _________________________________ . 3 . — O. MnO _________________________________ 2 . 13—0. 22 These basalts overlap the basalt of the Island Park caldera in silica content. However, the basalt of the Yellowstone Plateau is on the average less mafic than that of Island Park, having less magnesium and cal- cium and more silica, sodium, potassium, and titanium. Basalt and rhyolite liquids were erupted simul- taneously and mixed physically in two known in- stances in the Yellowstone Plateau (Wilcox, 1944; Boyd, 1961). Basalt containing rhyolite blebs grades into rhyolite containing basalt blebs, and as Boyd (1961, p. 403) concluded, “The character of this lava suggests mixing of rhyolite and basalt [liquids] at depth and extrusion as a single flow.” _ Olivine basalt must be extensive beneath the young rhyolite lava flows of the Yellowstone Plateau. South of West Yellowstone, large Bull Lake moraines were deposited by ice that flowed westward from the plateau. These moraines are younger than the rhyolite ash flows of the plateau but are overlain by the rhyolite lava of the Madison Plateau (Richmond and Hamilton, 1960). South from the latitude of West Yellowstone to the end of the exposed moraines, erratic blocks in the till are almost exclusively of uniform diabasic olivine ba- salt. Such rock is not present in appreciable quantity anywhere in the exposed source region of the ice, so the basalt must have come from the terrane buried by the post-Bull Lake rhyolite lava flows on the Central and Madison Plateaus. The large irregular caldera of the Yellowstone Plateau was probably flooded by uniform olivine basalt in pre-Bull Lake time, and during Bull Lake Glaciation the region now filled by the rhyolite of the Central and Madison Plateaus was probably a GEOLOGY AND PETROGENE‘SIS, ISLAND PARK CALDERA, EASTERN IDAHO C27 broad basalt plain rimmed by high—standing precollapse ash flows; the assemblage was analogous to that now present in the Island Park caldera. Rhyolite erupted subsequently has completely covered the basalt on the caldera floor. Pakiser and Baldwin (1961) found that the Yellow- stone Plateau has low Bouguer gravity and interpreted this to indicate that the plateau is probably underlain directly by a disk-shaped mass 2—4 miles thick of rhyolite or its plutonic equivalent that represents a thickened silicic crust, a collapsed magma chamber, or a roofless batholith. They assumed in making this inter- pretation that all pre-Pliocene rocks have the same den- sity and that no faults juxtapose pre—Pliocene rocks of different density. Other quite different interpretations can be made if it is assumed instead that pre-Pliocene structure is as complex beneath the plateau as in the surrounding highlands. The Buffalo Fork thrust, for example, carries relatively heavy Paleozoic strata and still heavier Precambrian crystalline rocks over the thick and very light Upper Cretaceous section (Love, 1956). The fault disappears under the rhyolite of the plateau south of Yellowstone Lake, and its continua- tion beneath the rhyolite could account for the gravity gradient, southeast of the lake, that Pakiser and Bald- win inferred to be due to great thickening of the rhyo- lite. The normal-fault blocks of the Teton and Madison Ranges project toward each other from opposite sides of the Yellowstone Plateau and may be continuous beneath the volcanic rocks; if they are continuous, then gravity interpretation is more complicated than is now realized. The possibility that large masses of magma still exist in the Yellowstone magma chamber makes the interpretation still more ambiguous. The Bouguer gravity intensity (—210 to —240 mgal) of the plateau is that to be expected in a region in isostatic equilibrium ‘ at this altitude (7,500—8,500 ft) and does not prove the causative light mass to be at the top of the crust. The rhyolite-and-basalt complex of the Yellowstone Plateau is larger than that of Island Park, but the two complexes may share a common history. In the Yellow- stone Plateau, a large central region collapsed while rhyolite ash flows were being erupted from a huge magma chamber. The resulting collapse basin was filled at some later stage by olivine basalt flows, and these in turn have since been buried by great flows of rhyolite lava. The Plateau complex of the Yellowstone has advanced one step further than has the complex of Island Park: in the Island Park caldera, the rhyolite lava flows and ash flows of the eastern rim have only partly buried the basalt that surfaces the caldera floor, whereas in the Yellowstone collapse depression the ba- salt has been hidden completely beneath younger rhyo- lite flows. I concluded previously (Hamilton, 1959) that the petrology of the rhyolite of the Yellowstone Plateau indicates a probable origin in the differentiation of ba- salt. Confirmation of this conclusion is afforded by the strontium-rubidium isotope studies of Yellowstone rocks by Pinson and others (1962) : the proportions of radiogenic strontium, common strontium, and rubidium oppose origin of the rhyolite by mobilization of silicic crustal materials and affirm its origin from material such as basalt. Because the rhyolite ash flows fill a broad topographic basin, I suggested that the volcanic rocks are in effect a gigantic lava flow rather than the crust of a roofless batholith as Daly (1911) suggested. The bulk of the hidden rock of this megaflow may be basaltic rather than rhyolitic: the rhyolite may be its differentiated scum, and the complex may be a lopolith. Probably most lopoliths are extrusive (Hamilton, 1960b). A stocklike chamber in which rhyolite magma overlay basalt magma may have reached the surface, and the rhyolite may have spread across the topographic basin; this theory is not greatly different from that proposed by Daly. SNAKE RIVER PLAIN The eastern half of the Snake River Plain, from the Island Park caldera to the vicinity of Twin Falls, is underlain by young olivine basalt flows that form a broad axial crest along the plain. The margins of the plain are alluvial lowlands within which flow the Snake and other rivers. Dipping under these lowlands from both north and south are Pliocene rhyolite ash flows and subordinate basalt lava flows (Stearns, Bryan, and Crandall, 1939; Mansfield, 1952; Malde and Powers, 1962). Some rhyolite vents were outside the present plain, but others were within it; and the inward dips of tuffs erupted from vents within the plain indicate tectonic tilting (Donald E. Trimble, oral commun., 1962). The silicic rocks thus far analyzed are from the margins of the plain at points more than 100 miles southwest of the Island Park caldera and are mostly within the following compositional ranges (Howard A. Powers and Donald E. Trimble, written commun., 1962) : Weight percent Weight percent —75 0. 2 — 0. 5 SiOz ___________ 68 TiOg ___________ A1203 __________ 11. 5 —13. 3 MnO __________ . 03— 0. 07 Iron as FeO____ 1. 5 — 3. 5 P205 ___________ . 03— 0. 09 MgO __________ . 01— 0.4 H20 ___________ .4 — 3 CaO ___________ .3—1.1 F _____________ .06— 0.11 NaZO __________ 2. 0 — 3. 5 C1 _____________ . 02— O. 12 K20 ___________ 5. 0 — 5. 8 These rocks generally contain less silica and more iron and magnesium than do the rhyolites of the Yellowstone Plateau and the Island Park caldera, but the assem- blages overlap in composition. C28 SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY The average and range of 38 unpublished analyses (Howard A. Powers, written commun., 1962) of olivine- augite basalt of Pliocene, Pleistocene, and Recent age in the central part of the Snake River Plain are as follows : Weight percent Average Range 46. 1 44 I ,_. o u wmowwawpmmo DO stflNONWHQm l I N) H,o __________________________________ 1C 0 According to Powers, there are compositional difier— ences between rocks of different ages in this analyzed suite. The older rocks are on the average less silicic, aluminous, and magnesian, and more titaniferous; they have more total iron, and a larger proportion of it is oxidized. The 'basalt of the Snake River Plain outside the Island Park caldera (table 1, No. 1) is well within these ranges for all components. The basalt inside the cal- dera (table 1, Nos. 2, 3), however, has more aluminum and calcium and less potassium, titanium, and phos— phorus than does nearly all the basalt of the central Snake River Plain; its composition more closely re- sembles that of the basalt of the Yellowstone Plateau. The high, northeast end—Island Park and the Yellow- stone Plateau—of the Snake River—Yellowstone vol- canic province differs compositionally from the rest, but the province throughout is one of bimodal olivine basalt and rhyolite. Rhyolite is also exposed along the axial crest of the eastern Snake River Plain. Three extrusive rhyolite domes rise above the plain between Idaho Falls and Arco, about 80 miles southwest of the Island Park cal- dera. Easternmost of the three is East Twin Butte, from which a specimen was collected by Howard A. Powers and analyzed by Dorothy F. Powers (U.S. Geo- logical Survey, Denver, Colo.) in 1958; the results of the analysis are as follows: Weight Weight percent percent 8102 _________________ 74. 89 ___________________ 0. 17 A1203 ________________ 12. 47 M110 ________________ . 05 Fe203 ________________ 2. 11 H20— _______________ 05 FeO _________________ . 20 H20+ _______________ 01 MgO ________________ . 07 02 _________________ 01 Geo _________________ . 59 N320 ________________ 3. 84 Subtotal _______ 99. 90 20 _________________ 5. 13 Less 0 for F____ . 07 TiOz _________________ . 18 P205 _________________ 02 Total __________ 99 83 This rhyolite is compositionally like that of the Island Park caldera and the Yellowstone Plateau. Despite the cover of olivine basalt throughout the Snake River Plain, there may be much rhyolite be- neath the central part. The rocks underlying the plain were tilted up during the extrusion of the westernmost and largest of the three rhyolite domes, Big Southern Butte, and were examined by W. Bradley Myers and me at the northeast flank of the dome. Of the 800 feet of volcanic rocks exposed, the upper 300 feet is inter- layered basalt and rhyolite; the lower 500 feet is basalt flows alone. The eastern Snake River Plain is a province of bi- modal volcanism of rhyolite and olivine basalt with little intermediate rock. Basalt dominates the surface exposures throughout most of the plain except in its up- per end, in the Island Park and Yellowstone region, but the characteristics of the rocks are similar throughout. COLUMBIA RIVER BASALT Washington (1922) applied the term “Oregonian basalts” to the basalt of the Snake River Plain, the Co- lumbia River Basalt of the Columbia Plateau to the northwest, and various Tertiary volcanic rocks of Ore- gon. The Columbia River Basalt is well known to be tholeiitic, and Washington’s misapprehension that the basalt of the Snake River Plain is part of the same tholeiitic province still persists. (See, for example, Turner and Verhoogen, 1960.) Buwalda (1923), Kirk- ham (1931), and others have recognized, however, that the Columbia River Basalt is older than and unrelated to the basalt of the Snake River Plain. The basalt of the Snake River Plain is uniformly olivine bearing and non- tholeiitic (Powers, 1960), whereas the Columbia River Basalt is everywhere tholeiitic (Waters, 1961). Basalts of the Snake River and Columbia River provinces are distinctly different in their occurrence, and rhyolite is not associated with the Columbia River Basalt. BIMODAL VOLCANIC PROVINCES In many volcanic provinces throughout the world, basalt and rhyolite have been erupted alternately or even simultaneously from interspersed vents. The genesis of these contrasting rock types must be closely related. Intermediate rocks——andesite and dacite—are subordinate or lacking in such bimodal provinces. Several Cenozoic volcanic fields in the Basin and Range province, south and southwest of the Snake River Plain, consist of rholite and quartz latite on the one hand and subordinate basalt on the other. In other bimodal provinces, basalt is dominant. Vol- canic terranes of early, middle, and late Precambrian age in northern Michigan include bimodal basalt-and- rhyolite assemblages. The lavas of the Keweenawan Series (late Precambrian) consist largely of basalt GEOLOGY AND PETROGENESIS, ISLAND PARK CALDERA, EASTERN IDAHO C29 (SiOz mostly between 46 and 50 weight percent) and subordinate rhyolite (SiOZ near 74 percent) ; inter- mediate types are of minor abundance (Broderick, 1935 ; Cornwall, 1951). Middle Precambrian metavolcanic rocks of Iron County are basalt and subordinate rhyolite (Gair and Wier, 1956). Early Precambrian metavol- canic rocks near Marquette are similarly bimodal, basalt being dominant over silicic rhyolite; these rocks are of “Keewatin” type and are the oldest rocks in this part of the Canadian Shield (Jacob E. Gair, oral commun., 1962). The processes that produce bimodal volcanism have operated throughout geologic time. The Tertiary volcanic province of western Scotland (Richey, 1948) consists of basalt, both tholeiitic and olivine bearing, and subordinate rhyolite and trachyte. The related plutonic complexes are of gabbro, and even ultramafic rocks, on the one hand, and of granophyre and granite on the other. Silicic rocks outbulk mafic ones at exposed levels in the plutonic complexes, but relative proportions vary widely among them. Silicic and mafic liquids were present simultaneously in each complex, and intermediate rocks are uncommon. Ero- sion of the Island Park and Yellowstone centers might reveal gabbro-and-granite plutonic centers similar to those exposed in Scotland. Basaltic and silicic lavas have generally been erupted alternately from the Iceland volcanoes Hekla and Askja during Recent time, and the siliceousness of the erup- tions is a direct function of the length of time between them (Thorarinsson, 1954). Askja simultaneously erupted rhyolite from its summit and basalt from its flank in 1875 (Spethmann, 1908) ; but in 1921—23 and in 1961, it erupted basalt from its summit (Thorarinsson and Sigvaldason, 1962). Differentiation of silicic mag- ma from basalt within single magma chambers seems required. An extensive Tertiary quartz latite tuff sheet contains bubbles of basaltic pumice throughout that total about 2 percent of the rock (Walker, 1962) ; simul- taneous eruption of liquid quartz latite and liquid basalt is required to explain this sheet and also other rhyolite- and-basalt flows and dikes (Gibson and Walker, 1962). Crustal structure of ,[celand is oceanic, not continental (Tryggvason, 1962) . All the basalt in some of these bimodal provinces, and much of it in all of them, is olivine basalt. By differen- tiation processes, only tholeiite might be expected to yield a rhyolitic residuum; perhaps the olivine basalt accompanying the rhyolite is itself a differentiate rather than a parent magma. Olivine basalt Of Olivine-basalt-only and of Olivine- basalt-and-trachyte provinces in general contains more alkalies, titanium, and phosphorus than do either most olivine basalt of the bimodal basalt-and-rhyolite prov- inces or nearly all tholeiite. It is perhaps because he was primarily concerned with the olivine basalt of a ' bimodal province (the British Tertiary) that W. Q. Kennedy (1933) concluded erroneously, in establishing the valuable concept of the contrast between olivine and tholeiitic basalts, that olivine basalt of all associations has a lesser content of alkalies than does tholeiite. LARGE CALDERAS . The Island Park caldera is larger than any other caldera having a regular circular or elliptical shape to which I have found reference. Numerous calderas as much as 15 miles across have many features like those of the Island Park caldera. Many irregular volcanic col— lapse depressions are larger. All or most very large calderas are at sites of volumi- nous eruptions of silicic ash flows. Among calderas as- sociated with more mafic or less silicic eruptions, Ngorongoro (12 miles in diameter) in Tanganyika is perhaps the largest known, and calderas larger than 6 miles in diameter are rare. Ash-flow calderas by con- trast commonly have diameters of 10—15 miles. Wil- liams (1941) and Matumoto (1963) described three such ash-flow calderas in Kyushu, Japan, each of which is the collapsed central elliptical part of a broad shield volcano built by the accumulation of thick silicic ash flows on an irregular foundation of more mafic volcanic rocks. Williams attributed the formation of these calderas, as of calderas in general, to collapse due to rapid extrusion of the supporting magma beneath them. Katsui (1963) demonstrated that this mechanism operated in Hok— kaido, Japan. The elliptical Valles caldera (Smith and others, 1961) , 12 by 15 miles, in north—central New Mexico resulted from the catastrophic eruption of about 50 cubic miles of rhyolite ash flows in a region of previous andesitic and basaltic volcanism. Following collapse, the caldera floor was domed, and flows of viscous rhyolite were erupted about a concentric ring within the caldera. The Creede caldera (Steven and Ratté, 1960) in southwestern Colorado is about 10 miles across and prob— ably originated as a result of eruptions of vast rhyolite ash flows and, later, quartz latite ash. The flat floor of the caldera was subsequently domed by intrusions, and viscous quartz latite formed local flows and domes around the margin of the caldera. The Timber Mountain caldera (Hinrichs and Orkild, 1961; Byers and others, 1964) Of southern Nevada is a complex caldera 15 by 18 miles that dropped from the center of a broad shield volcano composed of extensive thick rhyolite ash flows. Rhyolite flows and domes were extruded along the caldera scarp after collapse, and thin basalt flows were erupted locally. The Tibesti Mountains of the Sahara Desert consist of six major broad shield volcanoes of andesite and C30 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY basalt, each having a caldera 6—12 miles in diameter at its summit. The calderas are partly filled and the vol- canic shields are variably mantled by thick rhyolite ash flows that erupted from the calderas during collapse (Geze and others, 1959). Small flows of basalt have been erupted on the rhyolite Of most of the shields. Still larger but irregular collapse depressions have formed in ash-flow plateaus as a result of extremely voluminous eruptions. The Lake Toba depression in northwestern Sumatra is 60 by 20 miles in plan and is as much as 1,500 feet deep (Bemmelen, 1952). The large lakes in the ash—flow plateau of central North Is- land, New Zealand—the largest lake, Taupo, is 15 by 20 miles—occupy irregular structural collapse depres- sions that are defined by faults and downwarps pro- duced during ash-flow extrusion (Grange, 1937). An oval collapse depression 14 by 28 miles has been recog- nized in the volcanic San Juan Mountains of Colorado (Luedke and Burbank, 1961, 1962). The collapse de- pression of the Yellowstone Plateau (Boyd, 1961), 30 miles in diameter, is another irregular feature. All these calderas and larger irregular collapse de- pressions resulted from the very rapid extrusion of voluminous silicic ash flows. As some of the authors cited have emphasized, such calderas collapse when eruption is so rapid that the upper part of the magma chamber is emptied without being simultaneously re- filled by magma from beneath. Explosive discharge and collapse of frothy magma after overflow probably both contribute to caldera collapse. PETROGENESIS OF THE BASALT-RHYOLITE ASSOCIATION The Island Park caldera was apparently the scene of eruption of both rhyolite and basalt magmas from a. single magma chamber. The problem of the origin of these contrasting magmas is shared with the rest of the Snake River—Yellowstone province and with other bimodal rhyolite-and—basalt volcanic terranes. A popular hypothesis is that rhyolite Of this type (and all other types) forms by fusion of silicic rocks of the continental crust. Such conjecture does not ex- plain the intimate association of rhyolite and basalt liquids, is inapplicable to oceanic Iceland, and is not supported by the meager isotopic data; it therefore is not considered further here. An explanation is made instead in terms of liquid fractionation of basaltic magma into rhyolitic and basaltic magmas. Reasons for rejecting crystallization fractionation in this in- stance are given subsequently. Some bimodal volcanic provinces, including the Snake River Plain, contain broadly uniform olivine basalt as their dominant mafic rock type. Other prov- inces, such as the British Tertiary, include large amounts of both olivine and tholeiitic basalts as well as rhyolite, and some of these provinces also contain a variety of alkalic rocks in small volume. Much of the basalt associated with rhyolite in bimodal suites is oli- vine bearing and is markedly more mafic than tholeiite. The fractionation of tholeiite into olivine basalt and rhyolite is quantitatively adequate to account for the bimodal volcanic provinces, and a mechanism for such differentiation can be inferred. The compositional relationships are illustrated by figure 13. Plots of data representing tholeiite provinces scatter between those of data for the basalt and rhyolite of bimodal provinces; so the compositions of associated basalt and rhyolite can be accounted for in terms of a tholeiitic parent. None of the plots representing oxides in a tholeiite province falls exactly on the connecting line of any basalt-and-rhyolite province, but there is general concordance. The geometric relationships (the ratios of sides Of the similar triangles whose hypote- nuses are the line joining Olivine basalt, tholeiite, and rhyolite) indicate that if tholeiite does fractionate into olivine basalt plus rhyolite, between 5 and 30 percent Of the initial magmatic mass could become rhyolite, and the remainder, olivine basalt. The proportion varies with the provinces selected for comparison. Tholeiite is in general intermediate in composition between the basalts and rhyolites of the Snake River— Yellowstone subprovinces, with respect to the principal oxides and their ratios. This fact is illustrated by fig— ures 14 and 15, in which data are plotted for the olivine basalt and rhyolite of the Snake River—Yellowstone province and for the rocks of the adjacent Columbia River Basalt tholeiitic province. In some oxides or ratios the tholeiite and the olivine basalt overlap. The Columbia River Basalt is unusual among tholei— ites in that it has an uncommonly high content of iron and an uncommonly low content of magnesium (fig. 13; Powers, 1960). The total of iron and magnesium is, however, typical for tholeiite, and because it is a good index of variation in rocks, it is used as the abscissa in figure 14. In the iron-magnesium variation diagram, the plot representing the Columbia River Basalt can be seen to be on average near a straight line between olivine basalts and rhyolites of the Snake River—Yel- lowstone province for silica, calcium, sodium, and tita- nium. The relationships of aluminum and potassium are in the same direction but are not linear. Figure 15 illustrates the intermediate character of the tholeiitic Columbia River Basalt with respect to some other oxides. The relationship is linear on the potas- sium-calcium diagram and on the average linear but with scatter on the aluminum-calcium graph. There is still more scatter in the magnesium total-iron dia- g; GEOLOGY AND PETROGENESIS, ISLAND PARK CALDERA, EASTERN IDAHO C31 l “ l | l ‘ 15 H ~ h: A1203 ‘ EXPLANATION ‘ 10 b4 Basalt-and-rhyolite provinces # ~ \AU 0————O “\‘_Q> \T__ Yellowstone Island Park - C“‘\~.\‘\\\ Plateau Caldera ’ 5 _ S “‘:M\:\‘ . _____ . m . C’ Snake River Plain )- MgO I . O :5 10 k A r |_ EasFe ' :5 E FeO “2‘ 5 ~ 1 >. m :y 5 8 E 0 D. 5 E Lu“ 10 * i o f >—< 0 0210 5 — 7 i! ‘V > 8 " NaZO ’ 0 y 5 ~ _ *’ K20 l. l 0 l 5 r—i 'D- Tio2 :y 0 ._ b 40 80 I Si02, IN PERCENT BY WEIGHT 1? FIGURE 13.—Generalized silica-variation diagram for rhyolite and basalt of selected bimodal volcanic assemblages and for tho- }, leiitic basalt and diabase. The average compositions of associated rhyolite and basalt of three subprovinces of the Snake P River—Yellowstone province are connected by lines; data and their source are given in the text. Letters show average , compositions of tholeiitic basalt and of chilled margins of tholeiitic diabase sheets from various provinces, from these sources: Karroo diabase of South Africa (A) (Walker and Poldervaart, 1949, table 17, cols. 3-5) ; Jurassic diabase sheets r of McMurdo Sound region, Antarctica (M) (Gunn, 1962; Hamilton, unpub. data); diabase sheets of Tasmania (T) (Mc- Dougall, 1962, table 7); Triassic diabase of Pennsylvania and New Jersey (U) (Hotz, 1953, table 5, cols. 1—5; Columbia ,2) River basalt, Picture Gorge (C) and Yakima (C') Basalts plotted separately (Waters, 1961); and tholeiite (S, “Nonpor- phyritic central magma type”) of northwestern Scotland (Walker and Poldervaart, 1949). } C32 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY 80IIIIIIIIIIIIIITIIIIIIII jItvELLOWIONE ISLANDPARK 70— —- COLUMBIA OLIVINEBASALT _~5§ RIVER 0F SNAKE RIVER- _ E ‘g BASALT YELLOWSTONE N g a (THOLEIITE) PROVINCE gso_%§ r—Hr—A—\ _ ES _ E E Basalt of central - ‘5 g .1 Snake River Plain EE I (“r—M1 50—353 I'IO+ g — I II I I? e I E I — I Igl II I ' I I .I IgI ' I I 40 I + II II I I.E .I, _ I II I I8 E w ._ I ‘” 2 _ Ill+>§5 s 5‘15”“ I II a El E — N — .ggi’n'l (D A a I “gag I I | i —‘IZIELLOWSIONE I «I; I I I I " ISLAND PARK I ,I I I I I 1 | I ‘ I I I 10 I I II I I II I E 10‘ = 'I'>'_ # lg I I I _ mo II E II I .5 - 'IT'I I 1 a 5* I I: I ‘ A m: _ E _ «IEEI I: I _ o_ E OBI I8 I ,_ i I I I2 I — E _ >|_ I I ’5’?) E _ u; 0 3K I I I. I I IE ‘“ 9 3 ~K I I I Is I I I 1 , _ EON M l14,— x© ® _ z“ ”533i 'I ' _ I8 I 'I l _ o I I I II I I' I I I 54 II I II I — <2. : algal» : M — §E$T§ E _ _ II I ' I _ 0 I I I lf‘I“ k®. Q I I I III I i I N ‘ II ‘ IQCIJ ‘ E 2* TIJIXI I — I_ I I ,1 0 44 I | l' I' I i_I ' 1 I ' 'i II i ' i I _ I II I ><7 Location of analyzed specimen (Table 1) QUATERNARY fl 4 ., {pl/x 4) E Studies of the Zeolites GEOLOGICAL SURVEY PROFESSIONAL PAPER 504—D, E ‘6ng OF CAL/\ ‘ ‘ V P- MAR 151985 S. H ‘(I‘VOQ ”(PM IWOISSQIO-IJ Xmas 1991301099 SHLI’IOEIZ {'IHL J0 SflIflflLS—I‘HSOJ Studies of the Zeolites Composition of Zeolites of the Natrolite Group and Compositional Relations among Thomsonites Gonnardites, and Natrolites By MARGARET D. FOSTER SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY GEOLOGICAL SURVEY PROFESSIONAL PAPER 504—D, E UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1965 UNITED STATES DEPARTMENT OF THE INTERIOR STEWART L. UDALL, Secretary GEOLOGICAL SURVEY Thomas B. Nolan, Director The U.S. Geological Survey Library has cataloged this publication as follows: Foster, Margaret Dorothy, 1895— Studies of the zeolites. D. Composition of zeolites of the natrolite group. E. Compositional relations among thom- sonites, gonnardites, and natrolites. Washington, U.S. Govt. Print. Ofl’., 1965. v, 7; iii, 10 p. diagrs., tables. 30 cm. (U.S. Geological Survey. Professional paper 504—D, E) Shorter contributions to general geology. Each part also has separate title page. Includes bibliographies. (Continued on next card) Foster, Margaret Dorothy, 1895— Studies of the zeolites. 1965.” (Card 2) 1. Zeolites. I. Title. II. Title: Composition of zeolites of the natro— lite group. III. Title: Compositional relations among ,thomsonites, gonnardites, and natrolites. (Series) For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, DC. 20402 - Price 25 cents (paper cover) Studies of the Zeolites Composition of , Zeolites of the Natrolite Group By MARGARET D. FOSTER SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY GEOLOGICAL SURVEY PROFESSIONAL PAPER 504—D C/taracteristic variations in composition of natro/ites, sco/ecites, ana’ ineso/ites, indicated 5}! put/Med anaiyses CONTENTS Page Abstract ___________________________________________ D1 Mesolite ___________________________________________ Introduction _______________________________________ 1 Discussion _________________________________________ Natrolite __________________________________________ 2 References _________________________________________ Scolecite ___________________________________________ 3 ILLUSTRATION FIGURE 1. Relation between Ca( + Mg) and Na(+K) in natrolites, scolecites, and mesolites _____________________________ TABLES TABLES 1—4. Analyses and calculated atomic ratios of— 1. Natrolite _____________________________________________________________________________________ 2. Scolecite-_________‘___-___’ ____________________________________________________________________ 3. Mesolite ______________________________________________________________________________________ 4. High-Na mesolites ___________________________________________________________________________________ V Page D3 Page D6 Page D2 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY Studies of the Zeolites COMPOSITION OF ZEOLITES OF THE NATROLITE GROUP By MARGARET D. FOSTER ABSTRACT A study of the analytical data on members of the natrolite group of fibrous zeolites—natrolite, scolecite, and mesolite—— indicates that the scolecites are most constant in composition, with little substitution of Mg, Na, or K for Ca, and with little deviation from the theoretical value in the atomic ratios for [Ca, Al, and Si or in the molecular ratios for H20. In many natrolites the atomic ratio for Na, or even (Na+ K), is sig- nificantly less than theoretically required, and there is often not sufficient Ca present to compensate for the low Na. In some natrolites there is some evidence of replacement of Na, either by Ca or by CaAl; the CaAl also replaces an equivalent amount of Si. In others the deficiency in Na or (Na+K) cannot be accounted for in these ways. Water is generally close to the theoretical value, despite the low Na contents. There is some evidence that even small amounts of Ca tend to increase the H20 content. ‘ The mesolites vary widely in Ca and Na content, but three- fourths of the analyses studied are characterized by uniform and nearly theoretical Ca content and by variable Na content, which ranges from the theoretical to considerably lower than theoretical value. The nearly theoretical Ca, A1, and Si con- tents of these mesolites preclude either CaAl=NaSi or Ca:Na2 types of substitution to explain the low Na contents. \ The very high H20 molecular ratios of several of the low Na natrolites and mesolites may be due to adsorbed H20 or to hydronium ion whose presence would compensate for the low Na contents. INTRODUCTION Formulas for zeolites of the natrolite group—natro- lite, scolecite, and mesolite—suggest minerals of fixed composition. From his study of these minerals Hey (1932, 1933, 1936) concluded that this is true with respect to the ratio Si/Al, which is quite constant, and does not deviate notably from 1.5, but that there is more variation in their content of alkalies and alkaline earths. Most natrolites contain minor amounts of K and Ca; some scolecites contain minor amounts of Na and K; and many mesolites contain more or less Ca or Na than required by the formula. He found no evidence of NaSifiCaAl substitution, but attributed deviation in Na or Ca content to replacement of the Na2:Ca type. However, a casual inspection of the atomic ratios of the analyses he published does not corroborate these conclusions as to replacement. Structural studies have indicated a definite relation and close association between the cations and the water molecules. Detailed study of the natrolite structure by Meier (1960) has shown that each Na ion is surrounded by 6 oxygens (4 oxygen ions and 2 water molecules), and each H20 molecule is close to 2 Na ions. This relation is expressed in the 1:1 ratio between Na ions and H20 molecules. Scolecite is closely related crystal- lographically to natrolite and the chemical composition is similar except that the 2 Na ions and 2 H30 molecules in natrolite are replaced by 1 Ca ion and 3 H20 mole- cules in scolecite. The mesolite formula corresponds compositionally to 1 molecule of natrolite plus 2 molecules of scolecite, and the ratio of cations to water molecules is equivalent to 1:1 for the Na cations plus 1:3 for the Ca ions. Because of the close structural association between the cations and the water mole- cules, it would be expected that replacement of One cation by another or deficiency in cationic content would be reflected in H20 content. It seemed desirable, therefore, to reexamine analyses of these zeolites to learn the kind and extent of variation in composition, and the relation between variation in composition and H20 content. The. analyses used in this study were carefully selected from the literature, particularly with respect to age, summation of constituents, and summation of the atomic ratios of the tetrahedral constituents. Analyses included in the study were restricted to those made since 1900, except Where especially noted, for which the sum of the constituents was not less than 99.70, or more than 100.50, and for which the atomic ratios for Si and Al totaled 10 :t0.10. For purposes of comparsion all the atomic ratios for the analyses were calculated on the basis of 20 oxygen atoms. According to Hey’s (1955) formulas, calculations so based give quarter-cell atomic ratios for natrolite and scolecite, and twelfth cell ratios for mesolite. l D1 D2 SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY NATROLITE N34 .oAh .0815 .00 204.0 H30 The 27 analyses used to study the composition of natrolite are given in table 1, with their atomic ratios. The Na20 content of these analyses ranges from a high Of 16.73 percent tO a low Of 14.43 percent, or, in terms Of atomic ratios, from Na“, to Na3.57. K20, reported in about two-thirds of these analyses, ranges from 0.01 to 0.59 percent (0.10 K atomic ratio) except in N o. 26, which contains 1.27 percent K20 (0.20 K atomic ratio). In most of the analyses the atomic ratio for Na is less than the theoretical 4.00 indicated in the formula above. Even with K, the sum of the univalent cations (Na+K) is less than 4.00 in about three-fourths of the analyses, and is 3.85 or less in aboutvone-half of them. In several analyses (Na+K) is less than 3.70. CaO is reported in all but six of the analyses, but the amounts reported are generally low. Only six analyses reported more than 0.75 percent CaO (about TABLE 1.——Analyses of natrolite and their calculated atomic ratios [In order of decreasing N220 content] Composition (percent) Atomic ratio H20 Positive No. charge 810: A120; CaO NaaO K20 H20 Total Si Al Ca Na K Deter— Calcu- Differ- mined lated ence 1 ....................... 46 93 27.02 ........ 16.73 ________ 9.58 100.26 5.95 4.03 4.05 4.11 —0.06 4.11 2 ....................... 7 09 26. 99 ________ 16. 46 0 01 9. 80 100. 35 5. 97 4.03 4. 14 4.04 +. 10 4. 04 3 ....................... 47 1 27.1 ........ 16.4 ________ 9. 44 100.04 5. 96 4.04 3. 99 4.02 —. 03 4. 02 4 ....................... 47 40 26.88 0 05 16. 25 ll 9. 67 100. 36 6.00 4. 01 4.08 4. 02 +. 06 4. 00 5 ....................... 46 95 27. O6 27 15.97 ________ 9. 58 99. 83 5. 96 4.05 4.06 4. 05 +. 01 4. 01 6 _______________________ 46 72 26. 51 24 15.95 52 | 9.08 2 99.91 5. 97 3. 99 3. 79 4.18 -—. 40 4.13 7 ....................... 47 32 26. 30 50 15. 95 ........ 9. 50 99. 57 6. 02 3. 94 4. 04 4. 15 —. 11 4. 08 8 ....................... 47 17 26. 84 . 12 15.89 02 9. 58 8 99. 74 6.00 4. 02 4.06 3. 98 +. 08 3. 96 ....................... 47 22 27. 21 _...-_-_ 15.86 . 06 9. 70 100.05 5. 98 4.06 4.10 3. 90 +. 20 3. 90 10 ...................... 47 80 26.81 ........ 15.83 ........ 9. 69 100. 13 6.04 3. 99 4. 08 3. 88 +. 20 3. 88 11 ...................... 47. 69 27. 14 ________ 15 74 ________ 9. 56 100.13 6. 02 4.03 ________ 3.85 ........ 4. 02 3. 85 +. 17 3. 85 12 ...................... 46 91 27.10 .63 15 65 14 4 9. 72 100.15 5. 95 4.05 .08 3. 85 .02 3. 97 4.11 —. 14 4.03 13 ...................... 27. 36 . 15 63 . 13 9. 28 99.61 5. 90 4.10 . 11 3. 85 . 02 3. 94 4. 20 —. 26 4. 09 14 ...................... 47 33 27. 13 . 10 15 63 ........ 5 9. 67 99.8 6.00 4. 05 . 01 3. 84 ________ 4.03 3. 87 +. 16 3. 86 15 ...................... 46 53 26. 63 15 53 44 9. 62 5 100. 65 5.96 . 4.02 5 .08 3.85 .07 4.12 4.16 —. 04 4. 08 16 ...................... 47 45 27 40 07 15 45 42 9. 16 7 100.20 5. 98 4.07 . 01 3. 77 . 07 3. 86 3. 87 —. 01 3.86 17 ...................... 47 34 27 17 . 48 15 42 28 9. 47 8 100.17 5. 98 4.04 . 06 3. 77 . 04 3. 99 3. 99 . 00 3. 93 18 ...................... 47 60 27.40 13 15 36 23 9.47 100.19 6. 00 4.06 . 02 3. 74 .04 3. 98 3.84 +. 14 3. 82 19 ______________________ 47 15 27.39 30 15 27 50 9. 53 100.14 5. 96 4. 08 . 04 3. 74 .08 4. 02 3. 94 +. 08 3. 90 20 ...................... 47 38 27. w 54 14 96 41 9. 58 100. 50 5. 96 4.10 . 07 3. 65 . 06 4. 02 3. 92 +. 10 3. 85 21 ______________________ 46. 60 27. 21 Trace 14.80 31 10. 24 9 99.94 5. 98 4. 12 ________ 3. 68 . 05 4. 38 3. 73 +. 65 3. 73 22 ...................... 47. 33 27. 67 22 14.74 50 9. 64 100. 10 5. 98 4.12 . 03 3. 60 . 08 4.06 3. 77 +. 29 3. 74 23 ...................... 46. 6 27. 2 1 3 14. 7 < 01 9. 6 99. 4 5. 94 4. 09 . 18 3. 63 ________ 4.04 4.17 —. 13 3. 99 24. _ _. .................. 47. 29 27. 56 80 14. 63 59 9. 40 100.27 5. 96 4. 09 . 11 3. 57 . 10 3. 95 4. 00 ——. 05 3.89 25 ...................... 44. 85 27. 94 2 00 14.47 ________ 1'310. 68 11 100. 14 5. 78 4. 24 . as 3. 62 ________ 4. 37 4. 46 —. 09 4. 18 26 ...................... 47. 22 26 94 1 05 14.45 1 27 9. 28 100. 21 5. 98 4. 02 . 14 3. 54 . 20 3. 92 4. 16 —. 24 4. 02 27 ...................... 46. 42 27.27 1 04 14.43 . 47 9. 64 1‘ 99.86 5. 94 4.11 . 14 3. 57 .08 4. 11 4.07 +. 04 3. 93 I Includes 0.20 1120—. 2 Includes 0.20 F6203, 0.12 MgO (0.02 Mg atomic ratio), 0.08 P205, and 0.49 insoluble. 3 Includes 0.07 Fean and 0.05 MgO. 4 Includes 0.34 1120—. 5 Includes 0.13 H20—. 0 Includes 1.34 Fean and 0.12 MgO (0.02 atomic ratio). LOCALITY AND REFERENCE 1. Viagrande, Etna, Italy, Di Franco, S., 1929, Reale Accad. Lincei, Atti Cl. Sci. fis. mat. nat. Rend., ser. 6, v. 9, p. 660. Corporation quarry, Mount Royal, Canada, Harrington, B. J., 1905. Royal Soc. Canada Trans, v. 11, p. 25. Mouigit Elzgon, Uganda, Udluft, Hans, 1928, Arkiv Kemi, Mineralogi och Geologi, v. , p. . Grosspriesen, Bohemia, Tschermak, Gustav, 1917, Akad. Wiss. Wien Math.- Naturw. KL, Sitzungsber. Abt. 1, v. 126, p. 544. Bohemia, Niggli, Paul, 1923, Zeitschr. Kristallographle, v. 57, p. 656 analysis 9. . Kola7genin7sgsla, U.S.S.R., Kuz’menko, M. V., 1950, Akad. Nauk SSSiR Doklady, ~ v. . . Mori, Mount Baldo, Venetia Italy, Cavinato, Antonio, 1927, Realc Accad. Lincei, Cl. Sci. fls. mat. nat. Mam. ser. 6, v. 2, p. 325. Ice Valley Region, British Columbia, Canada, Phillips, A. H., 1916, Am. J our. 801., 4th ser., v. 42, p. 473. Kinbane (White Head), County Antrim, Ireland, nggli, Paul, 1923, Zeltschr. Kristallographie, v. 57, p. 656, analysis 10. 10. Brevik, Norway, Tschermak, Gustav, 1917, Akad. Wiss. Wien Math.-Naturw. KL, Sitzungsber. Abt. 1, v. 126, p. 544. 11. San Benito, Calm, Louderback, G. D., 1909, California Univ. Dept. Geol. Bull. 5, p. 331, analysis 23. 12. Pokolbln, New South Wales, Australia, Anderson, 0., 1904, Australian Mus- Recs., v. 5, p. 129. 13. Ben Lomond, New South Wales, Australia, Anderson, C., 1906, Australian Mus. Recs, v. 6, p. 420. 14. Wykertown, N.J., Milton, Charles, and Davidson, Norman, 1950, Am. Min- eral st, v. 35, p. 502. assesses 1 Includes 0.25 Fe203. 9 Includes 0.01 F6203. 0 Includes 0.52 FeaOa, 0.06 FeO, 0.05 MgO, and 0.15 T1027. 10 Includes 0.51 H20—. 11 Includes 0.20 Fean—FeO. 1' Includes 0.59 F6203. FOR ANALYSIS IN TABLE 1 15. l l q 2 . H 22. 23. 24. 25 . 26. 27. S” Thetiord mine eastern Quebec, Canada, Poitevin, Eugene, 1938, Toronto Univ. Studies, Geo . Ser., no. 41, p. 58. Valley of Chivrual River Lovozero massif, U.S.S.R., Vlasov, K. A., Kuz'menko, M. V., Es’kova, E. M., 1959, Akad. Nauk SSSR Inst. Mineralog. Geokhim, i Kristallokhim Redkika Elementov, p. 282, Analysis 3. . Amethyst Cove, Nova Scotia, Walker, T. L., and Parsons, A. L., 1922, Toronto 18. 19. 20. Univ. Studies, Geol. ser., no. 14, p. 64. Puy de Mannant, Puy-de—Déme, France, Hey, M. H., 1932, Mineralog. Mag. v. 23 p. 246, analysis 1. Cape Blomidon, King’s County, Nova Scotia, Hey, M. H., 1932, Mineralog. Mag. v. 23, p. 246, analysis 5. . Salesell, Ieitmeritz, Bohemia, Hey, M. H., 1932, Mmeralog. Mag. v. 23, p. 246, ana ys s 2. Vrahozily ( =Frauschile), nr. Boreslau., SE. of Teplits-Schonau, Bohemia, Nova- éek, Radirn, 1936, Praha, Nérodni Mus. Casopis, v. 110, p. 50. Rhiw, Carnarvonshire, Scotland, Hey, M. H., 1932, Mineralog. Mag. v. 23, p. 246, analysis 4. Red Island, Hawkes Bay, North Island, New Zealand, Mason, Brian, 1955, New Zealaud Jour. Sci. and Technology, sec. B, v. 36, RI 558. De Beers diamond mine, Kimberley, South Africa, Hey, . H., 1932, Mineralog. Mag., v. 23, p. 246, no.6. Highwood Mountains, Mont, Larsen, E. S., Hurlbut, C. S., Jr., Griggs, David, Bule, B. F., and Burgess, C. H., 1941, Geol. Soc. America Bull., v. 52, p. 1852. Snfilée Hilé, North Bergen, N. 1., Hey, M. H., 1932, Mineralog. Mag. v. 23, p. , no. . Budnany, S. W. of Praha, Bohemia, Kratochvfl, Frantisek, 1933, Praha N arod. niho Mus. Casopis, v. 107, p. 42. COMPOSITION OF ZEOLITES 0.10 Ga atomic ratio), and only one reported as much as 2.00 percent CaO (0.28 Ca atomic ratio). Several of the higher CaO values were reported in analyses significantly deficient in NagO, and thus raised the cationic content. However, little or no CaO was reported in several analyses that were significantly deficient in N a20(+K20). The Si and Al atomic ratios are generally 6.00 :|:0.10 and 4.00:];0.10, respectively. Only one analysis, No. 25, yields a Si atomic ratio with a greater deviation, —0.22. Deviations in Al content greater than 0.10 are slightly more common, being found in four analyses. The greatest deviation, +0.24 is in analysis N o. 25. The atomic ratios for Si, Al, Na, and Ca in N0. 25 indicate a slight CaAl replacement of NaSi. This type of replacement is also indicated in No. 13. Ca replacement of Naz(+K2) is indicated in No. 23, and No. 26. A little of both types may be suggested by the atomic ratios of Nos. 12 and 27. In most of the analyses, however, the atomic ratios for Ca are so low, and the Si and Al ratios are so close to the theoretical values that the low Na ratios cannot be explained by either type of replacement. The H20 content of these natrolites is very constant and very close (i0.15) to the theoretical 4.00 in 24 of the 27 analyses, even though (N a+K) is more than 0.20 deficient in one—third of the analyses. For some of the analyses that are low in Na but for which the atomic ratio of H20 is near 4.00 or even higher, the H20 value can be accounted for if the Ca present is considered as contributing to the H20 content in the 3:1 ratio as in scolecite. In No. 27, for example, the atomic ratio for (Na+K) is only 3.65, and consequently, accounts for a H20 molecular ratio of only 3.65 H20. However, the atomic ratio of Ca, 0.14, tripled (0.42) and added to 3.65 gives a calculated H20 ratio of 4.07, compared to the determined value of 4.11. Other analyses in which the atomic ratio for (Na+K) is much lower than the H20 molecular ratio and for which it is necessary to include the Ca value tripled to obtain a calculated H20 value comparable to the determined value are Nos. 17, 20, 24, and 25. This suggests that Ca is an intrinsic constituent of these natrolites. In other analyses, as in No. 26, inclusion of the Ca tripled in the calculated water computation yields a H20 value that is somewhat higher than the determined value, suggesting that some, at least, of the Ca may be extraneous to the natrolite molecule. In other analyses, as in No. 21, the determined water is considerably higher than the calculated. This analysis reports no Ca. Such a high H20 content may be due to adsorption of H20 in a humid atmosphere by the finely ground sample before analysis, H20— being seldom reported in analyses of zeolites, or a high H20 content may be 7‘43—604—64—2 D3 01“ THE NATROLITE GROUP due to the presence of hydronium ions, which would compensate for the low Na content. SCOLECITE 032.0A14.oSia.0020-6.0 H30 The 16 analyses used to study the composition and water content of scolecite are given in table 2, together with their atomic ratios calculated on the basis of 20 oxygen atoms. The CaO content of these analyses is quite constant, ranging only between 13.58 and 14.86 percent. The corresponding atomic ratio for Ca ranges only between 1.89 and 2.09, or 2.0:|:0.11. MgO is reported in only four of the analyses. The highest amount is only 0.32 percent or an atomic ratio for Mg of only 0.06. Na20 was reported in one-half of the analyses, most of which also reported K20. However, the amounts found were low, and the highest atomic ratio for (Na+K) in any of the analyses was only 0.20. For most of the analyses the atomic ratios for Si and A1 were within 0.10 of the theoretical 6.00 and 4.00, respectively. As these values and those for Ca are so close to the theoretical, there is little indication of replacement except in analyses 14, 15, and 16, in which the atomic ratios for Ca are the lowest and those for (N a+K) the highest. In N o. 15 the atomic ratio for (Na+K), 0.20, and the deficiency in the atomic ratios for Ca, 0.08, suggest slight replacement of Ca by (Na+K)2. The deviation in the atomic values of No. 15 for Si and A1 of —0.12 and +0.14, respectively, would suggest CaAl replacement for N aSi if the Ca were not already lower and the Na higher than the theoretical values. In Nos. 14 and 16 replacement is of the N 3.; ——90a type. The molecular ratios for H0 for all the analyses are within 0.20 of the theoretical value except for N o. 2, for which H20 is low, 5.71, and N o. 6, for which H20 is high, 6.32. However, the high H20 molecular value in No. 6 can be partially accounted for if H20 equivalent to Na is added to the tripled Ca value. Molecular values of H20, calculated on the basis of a 1 :3 relation between Ca atoms and H20 molecules and 1:1 relation between Na atoms and H20 molecules, agree fairly well with the determined H20 molecular ratios. For analyses 15 and 16, as well as for No. 6, it is necessary to include H20 equivalent to the (Na+K) present to produce a calculated H20 value more nearly approximating the determined H20 value. MESOLITE N81 .330 a1 .33A14 .oSio .0020-5- 33 H20 Analyses of 17 mesolites, together with their atomic ratios calculated on the basis of 20 oxygen atoms, are given in table 3. For 13 of the 17 analyses, Nos. 4—16, the atomic ratio for Ca is close to the theoretical value D4 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY TABLE 2.—Analyses of scolecite and their calculated atomic ratios [In order of decreasing CaO content] Composition (percent) Atomic ratio No. H20 SiOa A1203 CaO MgO N940 K20 H20 Total Si Al Ca Mg Na K De— Cal- Difler- termined culated once 5. 94 4.01 2. 09 5. 99 6. 31 —0. 32 5. 86 4.16 2.04 5. 71 6.12 -—. 41 5. 97 4. 02 2.04 5. 97 6. 12 -—. 15 5. 89 4. 13 2.03 5. 97 6. 09 —. 12 6. 11 3. 84 2.02 5. 80 6. 06 —. 26 5. 98 3. 98 2. 00 ________ . 13 02 6. 32 6. 15 +. 17 5. 98 4.02 1.98 ________ O3 ________ 5. 89 5. 97 —. 08 6. 06 3. 88 1.99 06 ........ . 00 6.05 6.15 —. 10 5. 95 4.04 1.97 04 09 .00 6.09 6.12 —. 03 5. 98 4.06 1. 95 ________________________ 5. 82 5. 85 —. 03 5. 94 4.11 1. 96 6.01 5. 98 +. 03 6.03 3. 97 1. 95 6. 05 5. 91 +. 14 6. 01 4. 02 1. 95 6. 10 5.85 +. 25 5. 97 4.03 1. 91 5.85 5. 93 --. 08 5. 88 4. 14 1.92 5. 94 5. 96 -—. 02 46. 37 6. 04 3. 96 1. 89 ________ . 16 02 6. 02 5. 85 —. 17 1 Ignition loss. 2 Includes 0.13 1120—. 3 Includes 0.55 F930;. 4 Includes 0.07 F8203. LOCALITY AND REFERENCE FOR ANALYSES IN TABLE 2 . Syhadree Mountains, Bombay, India, Hey, M. H., 1936, Mineralog. Mag. v. 24, p. 229 no. 1 B.M. no. 33887. Nr. Azhar, Caucasus, USSR, Shkabara, M. N., 1948, Akad. Nauk SSSR Doklady, v. 63, p. 730. . Teigarhorn, Berufjord, Iceland, Cavinato, Antonia, 1927, Reale Accad. Lincei, Cl. Sci. fls. mat. nat. Mem. set. 6, v. 2, p. . . Maderanerthal, Switzerland, Cavinato, Antonia, 1927, Reale Accad. Lincei. Cl. Sci. fis. mat. nat. Mem. ser. 6, v. 2, p. 331. Teigarhotn, Iceland, Koizume, Mitsue, 1953, Mineralog. Jour. Japan, v. 1, p. 39. . Valle di Viu, Valle di Lanzo, Italy, Gennaro, Virginia, 1929, Reale Accad. Sci. Torino Atti, v. 64 p. 141. . Poonah, India, Tschermak, Gustav, 1917, Akad. Wiss. Wien Math-naturw. Kl. Sitzungsber. Abt. I, v. 126,18I. . . An Gearna, Mull, Scotland, ’Lintock, W. F. P., 1915, Royal Soc. Edinburgh Trans, v. 51, p. 5. mqagnnhoapu of 1.33, being 1.3332007. In the other four, Ca is about 0.20 higher than theoretical value in three, (Nos. 1,2, and 3), and 0.12 lower in one, (No. 17). In the three analyses in which Ca is high, Na is low; in No. 17, in which Ca is low, Na is high. In this last analysis the relation between Ca and Na can be in- terpreted as replacement of 0.12 Ca by 0.22 Na, but the relations between Ca and Na and between Si and Al in Nos. 1, 2, and 3, do not permit such an interpretation. In the group of 13 in which Ca is very constant and close to the theoretical value, the amount of Na present varies considerably, and / in most is deficient. However, because of“ the fact that several contain a little K, or that Ca is slightly greater than the theoretical 1.33, the cationic charge is 4.00:l:0.10 for more than half of these analyses. In the analyses for which the positive charge is less than 3.90, slight adjustments in the amounts of Si and Al present produce the lower negative charges that just balance the lower positive charges. The Si ratio is 6.00:]: 0.10 for all the analyses given in table 3, and the Al ratio is 4.00:0.10 for all but two, Nos. 10 and 14, in which it is 4.11 and 4.16, respectively. However, as the respective Si ratios for these analyses are not low by similar amounts, replacement of the CaAl——>NaSi type is not indicated. Nor is similar 9. Alonéata, Eritrea, Italy, Scherillo, Antonio, 1938, Periodico Mineralogia, Rom a, v. p. . 10. Dry Lake, Clark County, Nev., Gianella, V. P., and Hedquist, Wilber, 1942, The Mineralogist, Portland, 0reg., v. 10, p. 108. 11. Teigarhorn, Iceland, Koizumi, Mitsue, 1953, Mineralog. Jour. Japan, v. 1, p. 39. 12. Teigarhorn, Iceland, Bauer, J aroslav, and Malkova, Ludmila, 1959, Sci. Papers Inst. Chem. Technology, Prague, p. 72. 13. Miage, Monte Bianoo, Italy, Cavinato, Antonio, 1927. Beale Accad. Lincei, Cl. Sci. fis. mat. nat. Mem., ser. 6, v. 2, p. 331. 14. Digby Gut, Annapolis County, Nova Scotia, Walker, T. L., and Parsons, A. L., 1922, Toronto Univ. Studies, Geol. Ser. no. 14, p. 68. ’ 15. Valle di V111, Valle di Lanzo, Italy, Gennaro, Virginia, 1929,Rea1e Accad. Sci. Torino Atti, v. 64, p. 141. 16. Bettolina Pass, Valle di Ayes, Monte Rosa, Italy, Gennaro, Virginia, 1929, Reale Accad. Sci. Torino Atti, v. 64, p. 137. replacement, or the reverse, except in a very slight degree, perhaps, indicated in any of the other analyses . because of the close agreement of the Si and Al ratios with the theoretical value. For most of the analyses the determined H20 values yield molecular ratios that agree fairly well with the the theoretical values. Furthermore, molecular H20 ratios calculated on the basis of a 1:1 relation between N a(+K) ions and H20 molecules, and a 1:3 relation between Ca ions and H20 molecules, agree within 0. 2O molecule with the H20 ratios based on the determined H20, except for analyses Nos. 9, 14, and 16. For Nos. 9 and 14, the determined H20 ratios are very high, 5.77 and 5.85, respectively. For N o. 16 the determined H20 ratio is only 5.42, quite close to the theoretical 5.33, but it is much higher than the calculated H20 because of the low Ca and Na content of this sample. Na is also low in Nos. 9 and 14. The high H2O content reported for these analyses may be due to hydronium ion, Whose presence would compensate for their low Na content. On the other hand, these high H20 values may be due to adsorption of H20 before analysis by the finely ground specimen in a humid environment. In most of these analyses only total water was reported: H20— and H20+ were differentiated only in analysis COMPOSITION OF ZEOLITES OF THE NATROLITE GROUP D5 TABLE 3.—Analyses of mesolite, and their calculated atomic ratios [In order of decreasing CaO content] Composition (percent) Atomic ratio Charge No. H20 5102 A120; CaO N320 K20 Hgo Total Si Al Ca Na K _ Positive Negative Deter- Calcu- Difier- mined lated once 26. 30 11. 15 4. 10 5. 90 4.07 1.57 1.04 5. 75 +0. 03 4. 18 4. 19 25. 63 10. 97 5. 02 5. 90 4. 01 1. 56 l. 29 5. 97 +‘. 20 4. 41 4. 37 26. 11 10. 90 4. 46 5. 95 4. 01 1. 52 1. 13 5. 69 +. 19 4. 17 4. 17 27. 00 10. 10 4. 98 5. 92 4. 10 1. 39 1. 24 5. 41 +. 08 4. 02 4. 02 26.02 10. 09 4. 50 6.06 3. 94 1. 39 1. 12 5. 31 —. 09 3. 92 3. 94 26.43 10. 06 4. 57 6.03 3. 99 1.38 1.13 5. 27 —. 16 3. 89 3. 91 26.84 10. 00 3. 82 5. 96 4.09 l. 39 . 96 5. 26 +. 05 3. 87 3. 89 26.66 9. 88 4. 66 5. 97 4.07 1. 37 1. 17 5. 31 +. 18 3. 94 3. 91 26. 64 9. 86 4. 20 5. 99 4.08 1. 37 l. 05 5. 16 +. 61 3. 79 3.80 27. 04 9. 73 4. 64 5. 96 4. 11 1. 34 1.16 5.18 +. 14 3.84 3. 83 26. 58 9. 72 4. 97 5. 99 4. 04 1. 34 1. 24 ________ 5. 5. 26 +. 02 3. 92 3. 92 26. 32 9. 72 5. 32 5. 96 4. 02 l. 35 1. 34 06 5. 39 5. 45 —. 06 4. 10 4. 10 25. 98 9. 69 4. 79 6. 01 4. 00 1. 36 1. 22 ........ 5. 25 5. 30 —. 05 3. 94 3. 96 27. 16 9. 67 4. 24 5. 94 4. 16 1. 35 1. 07 ________ 4 5. 85 5. 12 +. 73 3. 77 3. 76 26. 45 9. 35 5. 33 6. 01 4. 01 1. 29 1. 33 01 5. 39 5. 21 +. 18 3. 92 3. 93 26.86 8. 84 4. 60 6. 00 4. 09 5 1. 26 1.15 05 5. 42 4. 98 +. 44 3. 72 3. 73 26. 88 8. 77 6. 19 5. 95 4. 08 1. 21 1. 55 ........ 5. 23 5. 16 +. 07 3. 97 3. 96 1 Ignition loss, 13.53. 3 Includes 0.38 F6101 (Fefl- atomic ratio =0.04). 3 Includes 1.34 1120—. 4 Ignition loss, 13.36. 5 Includes 0.20 MgO (Mg atomic ratio=0.04). LOCALITY AND REFERENCES FOR ANALYSES IN TABLE 3 . Berufjord, Iceland, Cavinato, Antonio, 1927, Reale. Accad. LinceiICl. Sci. fis. mat. nat. Mem., Roma, ser. 6, v. 2, p. 339. . Kalageran, S. of Tiflis, USSR., Tvalchrelidze, A. A., 1922, Univ. Tiflis Bull. no. 2, p. 154. Mean of4 analyses. . Kilpatrick, Dumbartonshire, Scotland, Koizumi, Mitsue, 1953, Mineralog. Jour. Japan 17. 1, p. 39. Nishishioda-mura, Nagano Pret, Japan, Koizume, Mitsue, 1953, Mineralog. Jour. Japan, v. 1, . 39, no. 15. . Iceland, finely, M. ., 1933, Mineralog. Mag. v. 23, p. 423, no. 2. . Sgl31adre1e23 ounstains, Bombay, India, Hey, M. H., 1933, Mineralog. Mag, v. , p. ,no. . . Bhore Ghalzlg, Sylzgsdree Mountains, Bombay, India, Hey, M.I:1., 1933, Mineralog. ag. . . , no. . . Cape d’Or, ova Scotia, Walker, ’I‘. L., and Parsons, A. L., 1922, Toronto Univ. Studies, Geol. Ser., no. 14, p. 58, no. 1. mxraamgawnow 13. If the value for total water had been used to calculate the determined H20 molecular ratio for No. 13, the H20 ratio would have been 5.84, instead of 5.25, the value obtained when the H20+ value was used in the calculation. Thus H20 ratios based on total H20 values may be high because of adsorbed water. TABLE 4.—Analyses of high-Na mesolites PERCENT 8101 A120: 080 N30 K20 H20 Total 40. 03 27. 88 6. 08 10. 05 0. 40 1 11. 10 3 100. 29 41. 15 29. 49 5. 33 11. 02 . 25 13. 52 100. 76 40. 59 29. 69 5. 06 11. 00 . 51 13. 58 100. 43 ATOMS PER TWELFTH CELL El Al Ca(+Mg) Na(+K) H20 5. 48 4. 50 0. 92 2. 73 5. 07 5. 45 4. 60 . 76 2. 86 5. 97 5. 41 4. 66 . 72 2. 92 6. 03 I Includes 3.12 1120—. 1 Includes 0.85 F6203, 0.59 FeO, 0.22 MgO (0.04 atomic ratio), and 0.02 TiOz. 1. Rio Cambgne, Montiferro, Sardinia, Derlu, M. 1954, Periodico Mlneraiogia Roma, v. p. . .. 2. and 3. 'Kladno, Bohemia, Antonin, Rudolf, 1942, Kralovske Ceské Spolecnoste Nauk, Véstnik Art. 2, p. 11. 9. Berufjord, Iceland Cavinato, Antonio, 1927, Reale Accad. Lincei Cl. Sci. 63. mat. nat. Mem. oma, ser. 6, v. 2, 13./I . 10. Kviviig, Strdmii, Faroe Islands, Hey, . H., 1933, Mineralog. Mag. v. 23, p. 423, no. . 11. Faroe Islands, Giirgey, R., 1909, Tschermaks Mineralog. Petrog. Mitt. v. 28, p. 95. 12. Nova Scotia, Walker, T. L., and Parsons, A. L., 1922, Toronto Univ. Studies, Geol. Ser., no. 14, p. 58, no. II. 13. Osterb Faroe Islands Clarke, F. W., 1910, U.S. Geol. Survey Bull. 419 . 285, B. 14. Nishishloda—mura, Nagano Prat, Japan, Koizumi, Mitsue, 1953, ineralog. J our. Japan, v. 1, p. 39, no. 14. , 15. Poonah, Bombay, India, Bowman, H. L., 1909, Mineralog. Mag., v. 15 p. 220. 16. Yastreb, near Kurdjali, Bulgaria, Kostov, Ivan, 1958, Sofia Univ., icing.- Geolo .-Geo aph. Fakultet, Godishnik v. 53, p. 3 no. 4. 17. North able ountain, Golden, 0010., Clarke, F. W., 1910, U.S. Geol. Survey Bull. 419, p. 285. A. Three analyses of high Na mesolite are shown in table 4. These analyses are very similar to each other but very difierent from the mesolite analyses given in table 3, being respectivley lower in Si and Ca, and higher in Al and Na. Such compositions as indicated by these analyses cannot be derived from theoretical mesolite by either NaSiZCaAl or Na2<——>Ca replacement. They can only be derived from mesolite by replacement of an average of about 0.55 CaSi by 0.55 Nagsi. These analyses are very similar to some of the analyses of gonnardite published by Meixner, Hey, and Moss (1956) and by Kostov (1958), and to some high-Na thomsonites published or cited by Hey (1932). Such compositions can be derived from thomsonite by a combination of N az—eCa and N aSi——>CaAl replacement. DISCUSSION The most significant finding resulting from this study of analyses of the natrolite group of zeolites is that the amounts of Si, Al, and Ca present usually agree very well with the amounts required by the formulas but that the amount of Na present is often deficient. As a consequence, scolecites are the most uniform in com- D6 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY 2.4 Ideal scolecite 2.0 I 1.6 Ideal rpesollte O . 4 9 \ / ’E :. e * . m 0 O o E 1.2 o r— < (u . o Hugh—Na Rfiesolite 0.8 . .\ Ideal natrolite 0.4 0 0.4 0.8 1.6 2'0 2.4 2.8 3.2 Na ATOMIC RATIO FIGURE 1.—Relation between Ca(+ Mg) and Na(+ K) in natrolites, scolecites, and mesolites. position and conform most closely with the require- ments of the formula, and natrolites deviate most from the requirements of the formula. In mesolites the Ca is usually close to the theoretical value, but the Na is often significantly lower than the theoretical value. As the other constituents deviate so little from the theoretical, the deficiency in Na cannot be interpreted, generally, in terms of either type of replacement, Ca—eNag or CaAl—eNaSi. In some natrolites there is sufficient Ca or K present to produce a positive cationic charge close to the theoretical 4.00. In some mesolites also the Ca, although close to the theoretical, is enough higher to bring the positive chargerclose to 4.00. However, in other natrolites and mesolites the positive cationic charge is only 3.75, or even less. Slight adjustments in the amounts of Si and Al present produce the lower negative charges that just balance the lower positive charges. The relation between Ca and Na in these zeolites is shown graphically in figure 1. The points representing the relation between Ca and Na in scolecites all fall in a very small area closely grouped around the point representing ideal scolecite, Whereas the points repre- senting the relation between Ca and Na in mesolites and natrolites are more scattered. Most of the points representing the relation between Ca and Na in analyses of mesolites and natrolites fall left of the points repre- senting ideal mesolite and ideal natrolite, respectively, thus indicating lower N a than ideally presumed to be present. However, three points representing the rela- tion between Ca and Na in samples of high Na mesolite fall lower and far to the right of the mesolite area; closer to the natrolite area than to the mesolite area. In composition these analyses resemble gonnardites or high-Na thomsonites rather than mesolites. Most H20 molecular ratios based on reported H20 contents are close to the theoretical value specified in the formula. Also, most of the H20 values calculated from the Ca(+Mg) and Na(+K) present agree Within 0.20 molecule of the H20 values based on the amount of H20 reported in the analysis. In general the deter- mined H20 ratios are higher than the calculated ratios. A few determined H20 values are very high. As most zeolite analyses report only total H20, and do not diiferentiate H20— and H20+, these higher values for determined H20 may be due to adsorbed H20, which is probably very slight in most of the samples analyzed but in some may be considerable, and so produce an unrealistically high H20 content particularly if the finely powdered sample has been in a humid environ- ment before analysis. The dehydration curves of Koizumi (1953) and Peng (1955) indicate that H20 COWOSITION 0F ZEOLITES that is an intrinsic part of the structure does not begin to come off below about 150° C; consequently deter- mination of H20— at 110° C is perfectly feasible. It is therefore recommended that H20— be routinely determined in the zeolites. The highest H20 values in both the natrolites and the mesolites occur in analysis that are very low in Na and that have the lowest positive charges. It may be, therefore, that these high H20 values are caused by hydronium ions which compensate for the lOW cationic content of the zeolite. REFERENCES Hey, M. H., 1932, Studies on the zeolites. Part III. Natrolite and metanatrolite: Mineralog. Mag., v. 23, p. 243-289. 1933, Studies on the zeolites. Part V. Mesolite: Minera- log. Mag., v. 23, p. 421—447. or THE NATROLITE GROUP D7 1936, Studies on the zeolites. Part IX, Scolecite and metascolecite: Mineralog. Mag, v. 24, p. 227—253. 1955, An index of mineral species and varieties arranged chemically, 2d ed.: London, British Museum (Natural History), 728 p. Koizumi, Mitsue, 1953, Studies on water in minerals. Part I. The difl’erential thermal analysis curves and the dehydration curves of zeolites: Mineralog. Jour. Japan, v. 1, p. 36—47. Kostov, Ivan, 1958, Zeolites in Bulgaria: scolecite, mesolite, “gonnardite,” and thomsonite: Sofia Univ. Biolog.— Geolog.—Geograph. Fakultet, Godishnik, v. 53, p. 1—24. Meier, W. M., 1960, The crystal structure of natrolite: Zeitschr. Kristallographie, v. 113, p. 430—444. Meixner, Heinz, Hey, M. H., and Moss, A. A., 1956. Some new occurrences of gonnardite: Mineralog. Mag., v. 31, p. 265— 271. _ Peng, C. J ., 1955, Thermal analysis study of the natrolite group: Am. Mineralogist, v. 40, p. 834—856. Taylor, W. H., Meek, C. A., and Jackson, W. W., 1933, The structures of the fibrous zeolites: Zeitschr. Kristallographie, V. 84, p. 373—398. Studies of the Zeolites Compositional Relations among Thomsonites Gonnardites and Natrolites By MARGARET D. FOSTER SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY GEOLOGICAL SURVEY PROFESSIONAL PAPER 504—E Relations among tfiomsom’tex, fizg/z-soa'z'um Mom— :om'tes, gonnara’z’tes, flzg/z—soa’z'um mesa/ins, and natrolz’z‘es as indicated 5} puélz’sfiea’ anaZyse: CONTENTS Page Page Abstract ___________________________________________ E1 Formulation _______________________________________ E9 Introduction _______________________________________ 1 Other lines of replacement ___________________________ 9 Thomsonite ________________________________________ 2 Conclusion _________________________________________ 10 Gonnardite ________________________________________ ‘ 4 References _________________________________________ 10 A thomsonite-gonnardite—natrolite isomorphous series?--- 6 ILLUSTRAITONS Page FIGURE 1. Relation between Ca and Na in thomsonites _____________________________________________________________ E3 2. Relation between Ca and H20 in thomsonites ____________________________________________________________ 4 3. Relation between Ca and Na in gonnardites, and high Na thomsonites and mesolites ________________________ 5 4. Relation between A1 atoms per unit cell and mean index of refraction in thomsonites _________________________ 7 5. Relation between Na atoms per unit cell and mean index of refraction in thomsonites ________________________ 7 TABLES, Page TABLE 1. Atomic ratios calculated from analyses of thomsonite reported by Hey (1932) ________________________________ E2 2. Atomic ratios calculated from selected post-1870 analyses of thomsonite cited by Hey (1932) ___________________ 2 3. Analyses of gonnardites and high-Na mesolites, and their atomic ratios _____________________________________ 4 III SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY Studies of the Zeolites COMPOSITIONAL RELATIONS AMONG THOMSONITES, GONNARDITES, AND NATROLITES By MARGARET D. FOSTER ABSTRACT A study of post—1885 analyses of thomsonite selected from Hey’s 1932 compilation confirms the isomorphous series between thomsonite and faroelite. This isomorphous series, which is characterized by replacement of CaAl by NaSi, was recognized by Winchell (1925, 1926) and Hey (1932). Hey’s formula for gonnardite is related to thomsonite in the same way and carries this line of replacement a step beyond faroelite. How- ever, the total number of cations, and the relation between Si and Al and Ca and Na in analyses of gonnardite, including those published by Meixner, Hey, and Moss (1956) indicate re- placement of Ca by N32 as well as replacement of Ca as CaAl by NaSi. Similar numbers of total cations, and similar rela- tions between Si and A] and Ca and Na were also found in some analyses of high Na thomsonites and in high Na mesolites. Plotted points representing the relation between Ca and Na in these analyses fall, not along the thomsonite-faroelite line, but along a line representing dual replacement, to the same degree, of Ca by Na, and of CaAl by NaSi. Such dual replace- ment of Ca in thomsonite leads theoretically to natrolite: Thomsonite: C32,oNal.oA15.osl5.oOm-6.0 H20 031.0N82.5A14.53i5.5020'5-0 H20 Natrolite: Ca.oNa4_oAl4,oSis.002o-4.0 H20 The analyses alone suggest an isomorphous series between thomsonite and natrolite, with gonnardite as the intermediate member. For such a series only formulas indicating ranges of composi- tion are adequate. As there appears to be no natural hiatus in composition between thomsonite and gonnardite, the di- vision was arbitrarily fixed at the point where Ca and Na are equal. The suggested range formulas for the series are: Thomsonite: Nal _o_1450812.0—1.uAl5,o—4,ssl5_o—5_202o-6.0—5.6 H20 Gonnardite: N31.5-33084.a_o.4Al4.3—43SI53—5,3020-5.6—4.4 H20 Natrolite: Nam—4.0030.4—o.oA14.2—4.oSi5.s—o.oozo'4-4-4-0 H20 Although thomsonite and natrolite have similar alumino— silicate frameworks, they differ considerably in detail. Thus there may be a structural hiatus between thomsonite and natro- lite like that between muscovite and lepidolite. No study has been made of the structure of gonnardite. The indices of refraction of gonnardite are intermediate between those of thomsonite and natrolite, those of thomsonite the higher, and those of natrolite the lower. However, the optic sign of gon- nardite is negative, and those of thomsonite and of natrolite are positive. INTRODUCTION The formulas for thomsonite and gonnardite, like those for natrolite, mesolite, and scolecite, suggest minerals of fixed and definite composition, although they are well known to be quite variable in composition. Winchell (1925) recognized that there is isomorphous N aSifiCaAl replacement in thomsonite, as between thomsonite, CamNal,0A15,oSi5.0020-6.0 H20 and Cal_25Na,.75Al4.258i5_75020.5.0 H2O. Hey (1932) concurred with this view but considered that there is also con- siderable N 32:03: replacement, which Winchell had considered unimportant. However Hey’s formula for gonnardite, 2[NazCaAl4Si6020-7H20], is related to thom- sonite by NaSi—>CaAl replacement only, although atomic ratios derived from analyses of gonnardite, including those published by Meixner, Hey, and Moss (1956), indicate both N aSieCaAl and N age Ca substi- tution. The relation between H20 and the cations in thom- sonite is less simple than in natrolite or mesolite. Taylor, Meek, and Jackson (1933) assume the environ- ment of the Na ion and one of the Ca ions to be similar to that of the Na in natrolite, except for an additional H20 molecule that increases the coordination from six— fold to sevenfold with 4 oxygen ions and 3 water mole- cules. The other Ca ion they assume to be in eight- fold coordination, with 6 oxygen ions and 2 water molecules. The relation between water molecules and cations is as though there were 1 water molecule for each Na ion, 2 for one-half of the Ca ions and 3 for the other half, or an average of 2.5 to 1 for all the Ca ions. The water content of gonnardite seems to be in doubt. Hey ,(1955) gives the .formula as2[ N3203A14Si6020'7 H20], but in 1956 Meixner, Hey, and Moss give the formula as [(Ca,Na)6_8(Si,Al)2OO4o-12 H20], and Deer, Howie, and Zussman (1963 p. 359) give the formula as NagCa[(Al,Si)5Om]2-6 H20. These amounts of water seem high in a fibrous zeolite in which Na is the principal cation. In other fibrous zeolites in which Na is a El E2 principal cationwnatrolite and mesolite——the ratio of H20 molecules to Na ions is 1:1, compared to the ratio of 3:1 or 2.5:1 for Ca-dominant zeolites like scolecite and thomsonite. As Na is the predominant cation in gonnardite, H20 would be expected to be lower in gonnardite than in thomsonite, rather than the same or higher as indicated by these formulas. It is significant that in the analyses of gonnardite given by Meixner, Hey, and Moss, the two having the highest Na content yield molecular ratios for H20 of only 5.14 and 5.19. The purpose of this investigation is to study the degree and kind of variation in composition that occurs in the thomsonites and gonnardites, and the relation of these minerals to each other, and to examine the prob- lem of the water content of gonnardite. The analyses of thomsonite used were selected from those reported or cited by Hey (1932), using only analyses published after 1885, in which the sum of the tetrahedral atoms was 10.0 i010. In addition to the analyses of gonnar— dite published by Meixner, Hey, and Moss, several other recent analyses from the literature were also used. Atomic ratios were calculated from the analyses on the basis of 20 oxygen atoms, and represent, according to Hey’s (1955, p. 166) formulas, quarter-cell values for thomsonites and half-cell values for gonnardites. THOMSONITE N81.oCaa.oAla,oSis.003o-6.0 H30 Atomic ratios for selected post-1885 analyses of thomsonite reported or cited by Hey (1932) are given in tables 1 and 2 together with calculated values for H20. The atomic ratios show wide ranges in the values for Si and Al. The ranges for Ca and Na are even greater, from 2.14 to 1.00 for Ca, from 0.74 to 2.71 for Na, and from 2.58 to 3.71 for (Ca+Na). However, Ca values greater than 2.00 are found in only four analyses and these values are 2.02, 2.04, 2.07, 2.14. Values for Na less than 1.00, were found in only seven; of these seven, only three, 0.89, 0.88, and 0.74, are less than 0.90. Both types of replacement recognized in thomsonite affect the Ca and Na content, consequently the kind and amount of replacement can be estimated only by other deviations in composition. A Si value greater than 5.00, the theoretical amount according to the above formula for thomsonite, together with an Al value deficient by about the same amount, indicates replace- ment of an equivalent amount of Ca by Na. A positive difference in the expression [(Ca+Na)+3.00] indicates that this amount of Ca has been replaced by Nag; a negative result indicates the amount of Ca replacing Nag. The fact that in the'ranges in Ca and Na quoted above most Ca values are less than 2.0, with few signifi- cantly above, and that most Na values are greater than 1.00, with few significantly less, indicates:that replace- SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY TABLE 1.——Atomtc ratios calculated from analyses of thomsonite reported by Hey, 1932 [Numbers refer to Hey, 1932, table 1, p. 54-55] 1110 N o. Oa(+Mg) Na(+K) (Oa+Na) A1 81 , Deter- Calcu- Difier mined lated once 1 _____ 1. 62 1. 48 3. 10 4. 52 5. 43 6. 34 5. 53 +0. 81 2 _____ 1. 67 1. 20 2. 87 4. 57 5. 42 5. 58 5. 38 +. 20 3 _____ 1. 84 . 74 2. 58 4. 58 5. 46 6. 10 5. 34 +. 76 4 ..... 1.71 .89 2 60 4. 62 5. 46 6. 44 5. 17 +1.27 5 ..... l. 00 2. 71 3. 71 4. 62 5. 36 5. 48 5. 21 +. 27 6 ..... 1. 41 1. 70 3. 11 4. 62 5. 40 6. 44 5. 22 +1.22 7 ..... 1. 72 1. 18 2. 90 4. 65 5. 36 5. 50 5. 48 +. 02 8 _____ 1. 76 1. 24 3. 00 4. 68 5. 29 6. 12 5. 64 +. 48 9 _____ 1. 45 1. 71 3. 16 4. 70 5. 32 5. 46 5. 33 +. 13 10.- - 1.82 1. 24 3.06 4. 73 5. 23 5. 92 5. 79 +. 13 11. . 1. 74 1. 26 3. 00 4. 78 5.23 5.89 5. 61 ' +. 28 12. - 1. 98 1. 22 3. 20 4. 87 5. 06 5. 82 6.17 —. 35 13.. - 2. 08 . 88 2. 96 4. 90 5. 06 6. 10 6. 08 +. 02 14.. _ 1. 98 . 97 2. 95 4. 93 5. 07 5. 61 5. 92 —. 31 15-. 2. 00 1.04 3. 04 5.03 4. 97 5. 95 6.04 —. 09 16--.. 2. 02 . 94 2. 96 5.04 4. 98 6.17 5. 99 +. 18 TABLE 2.—-Atomic ratios calculated from selected post-1870 analyses of thomsonite cited by Hey, 1932 [Numbers refer to Hey, 1932, table 2, p. 58-64] H10 No. Ca(+Mg) Na(+K) f(Ca+Na) A] St Deter- Calcu- Difler mined lated ence 243... 1. 59 1.58 3.17 4. 76 5 24 5. 52 5. 56 -0. 04 29.... 1. 56 I. 30 2. 86 4. 64 5. 40 5.42 5. 20 +. 22 32.. - 1. 94 1.04 2 98 4. 98 5.04 6. 23 5. 89 +. 34 33.. - 1.98 1.10 3.08 5.06 4.94 6.00 6.05 —.05 45-- - 1.96 1.04 3.00 5. 03 4.98 6. 08 5. 94 +—. 14 47.--- 1. 92 1. 16 3.08 4. 78 5. 16 6.03 5. 96 +. 07 49. - 2 04 1. 01 3.05 5. 02 4. 91 6.05 6. l1 —. 06 54-- 1. 84 l. 10 2. 94 4. 80 5. 20 5. 98 5. 70 +. 28 58-- _ 1. 94 1. 10 3.04 4. 95 5.05 5. 81 5. 95 —. 14 64.- - 1.94 . 91 2. 85 4. 93 5.11 5. 73 5. 76 —. 03 66.- 1. 18 2 18 3. 36 4. 67 5. 37 6.54 5. 13 +1. 41 69. _. 1. 58 1. 62 3. 20 4. 72 6. 26 5. 08 5. 57 +. 49 71.-.. 1.64 1. 30 2 94 4. 57 5. 43 5. 84 5. 40 +. 44 72--_ 1. 68 1.06 2 74 4. 78 5. 22 5.64 5. 26 +. 38 76-. - 1. 73 1.06 2. 79 4. 83 5 24 5. 71 5. 38 +. 33 78...- 1. 68 1. 21 2.89 4. 62 5. 40 5. 68 5. 41 +. 27 79. - 1. 68 1. 14 2. 82 4. 68 5. 36 5. 66 5. 34 +. 32 80.- 1. 64 1.24 2. 78 4. 50 5. 56 5. 34 5. 09 +. 25 81. - 1. 69 1. 43 3. 03 4. 60 5. 39 5. 68 5. 43 +. 25 83. .. 1. 59 1. 36 2. 95 4. 60 5. 42 5. 77 5. 34 +. 43 92a-- 2. 14 . 95 3. 09 5. 02 4. 92 6. 24 6. 30 —. 06 94.... 1.97 1. 04 3. 01 4. 98 5. 02 5. 90 5. 96 -—. 06 95. ... 1. 74 1. 05 2. 79 4. 96 5. 02 6. 79 5. 40 +. 39 100... 1.64 1. 50 3. 14 4. 61 5. 36 5. 59 5. 60 —. 01 ment of Ca by Na, by one method or the other, greatly predominates over the reverse and that replacement of Na by Ca is of relatively little importance in the thomsonites. The atomic ratios yielded by some analyses of thom- sonites are so close to the theoretical values that little replacement of either type is indicated. For example, in N o. 15, table 1, none of the atomic ratios deviate more than 0.04 from the theoretical values, and the molecular value for H20 is 5.95 as compared with 6.00. Other analyses whose atomic ratios, given in tables 1 and 2, are very close to the theoretical values, are Nos. 14, 16, 32, 33, 45, 49, and 94. The Si, Al, and (Ca+Na) values for all the other analyses whose atomic ratios are given in tables 1 and 2 indicate both types of replace- COMPOSITIONAL RELATIONS AMONG THOMSONITES, ment, except for N 0s. 8, 11, and 81. In these replace- ment of CaAl by NaSi is indicated; the (Ca+Na) values of 3.00 and 3.03 indicating that no Na2:‘Ca replacement has occurred. In almost all the analyses in which both types of replacement are indicated, the Si value is usually greater than 5.00 and the Al value is less, indicating that replacement of CaAl by NaSi is more common in thomsonites than replacement of NaSi by CaAl. With respect to N a2:‘—Ca replacement, 17 of the 40 (Ca+N a) values in tables 1 and 2 are greater than 3.00, indicatng replacement of Ca by N32, and 20 are less than 3.00, indicating replacement of N32 by Ca. As in most thomsonites replacements of the N aSi:CaAl type are of N aSi for CaAl. Thus in some there is dual replacement by Na, but in others there is both replace- ment of CaAl by NaSi and replacement of Na; by Ca, with the replacements offsetting each other to greater or less degree. Although the thomsonites whose atomic ratios are given in tables 1 and 2 differ widely in content of Ca and Na, the sum of the atomic ratios for Ca and Na, (Ca+Na) is quite constant, 3.00:1;020 in all but eight of the analyses (Nos. 3, 4, 5, 66, 72, 76, 80, and 95). In Nos. 76 and 80 the difference is only 0.21 and 0.22, respectively. The relation between Ca and Na in thomsonites is shown graphically in figure 1. The points representing thomsonites whose atomic ratios for Ca and Na are close to the theoretical 2.00 and 1.00, respectively, cluster closely around the asterisk that represents theoretical thomsonite. A few points fall a very short way along the line representing Nag—90a replacement, but more fall along or below the line representing N aSi replacement of CaAl. However, none fall as far along this line as the asterisk representing theoretical faroelite. Even the point representing the average Ca and Na ratios for the five Faroe Islands zeolites cited by Hey (1932)—Nos. 19, 21, 22, 23, and 24a—fall short of this point. Farther out along this line is the point representing the “gonnardite” calculated by Hey (1932) from an analysis of a spherulite made up of thomsonite and gonnardite. The points that fall below the NaSi—>CaAl replacement line represent \thomsonites in which replacement of CaAl by NaSi is accompanied by some replacement of Nag by Ca. Most of the points representing the Table Mountain, 0010., thomsonites (Nos. 78, 79, 80, 81, and 83, table 2), fall in this area. The points representing the relation between Ca and Na in thomsonites that have the highest Na content (Nos. 5, 9, 24a, 66, and 69) fall along a line between the Nay—90a replacement and the NaSi——>CaAl replacement lines. This middle line represents dual replacementzreplacement of Ca by N a2 and of Ca as GONNARDITES, AND NATROLITES E3 4.0 3.6 3.2 4/ 81s;- 2.8L_.>\ g 2.4 ’2 a: 9 20 \r g ' D(IV)Hey'sgonnardit'e"\ D. \ }_ . j: Winchell's enci member A\6.9. \ z 1.6 I \69 7‘43 Thepretical faroelitef o ‘\ D (Ill) Hey's high-Si thomsonite/ikg \\ l \ 1'2 Average faroelité 0' k ' Theoretical 0 ,thoms‘anite O O . 0.8 . \\ \ \ \ \\ 0.4 ‘ \\ \ \ \ \ ‘\ 0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 Ca ATOMIC RATIO FIGURE l.—Re1ation between Ca and Na in thomsonites. CaAl by NaSi to the same degree. The highest Na ratio in tables 1 and 2 is that in analysis 5, but the highest Na ratio in the analyses cited by Hey (1932) is in analysis 28. The value he gives, 12.50, is for the unit cell, which would be 3.125 in terms of the quarter cell as used in this study. In the same terms the ratio for Ca is only 0.68. The point representing the relation between Ca and Na in this analysis falls considerably farther along the dual replacement line than does the point representing N o. 5 which actually falls between the dual replacement line and the Nag—60a line. Analysis 28 is cited by Hey as that of a thomsonite. although it had been earlier considered to be that of a mesolite. It was not included in table 2 because of its age (1836). The relation between Ca and H20 in thomsonites is shown graphically in figure 2. Although thomsonites that have the same Ca ratio may differ greatly in H20 content, the points in figure 2 do show a general down- ward trend in H20 content with decrease in Ca. In general, analyses having high Ca contents tend to have high H20 contents, and analyses having low Ca con- tents tend to have low H20 contents, except for analyses 1, 4, 6, and 66, in which H20 is very high. The points representing these analyses fall completely outside the pattern made by the other points. E4 6.60 66. x6(H) 4(H)x 6.40 x1(H) 6 0 . / .2 EXPLANATION (H) ‘<(H) x x(H) x(H) 9 x(H) o L: 6.00 — Analyses reported by Hey . 6: (HM 6: ° (H)x (H) ' g Analyses cited by Hey o / (|-.l)x 8 5.80 e Theorietical .7 . ' tl'omsonite g l . q; 5.60 /-x “(m I (H) x504) XU‘y‘ (H)x 5.40 ° / . 5.20 . 5.00 0.80 1.00 1.20 1.40 1.60 1.80 2.00 2.20 Ca ATOMIC RATIO FIGURE 2.——Relation between Ca and H20 in thomsonites. GONNARDITE Nazocai.oA14.oSio.0020'7 H20 Five analyses of gonnardite were published by Meixner, Hey, and Moss (1956), including two (D and SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY E) that Hey had published in 1932 as an analyses of gonnardite-thomsonite spherules. The outer part of the spherules was described as thomsonite, and the inner part as gonnardite. When the analyses were repub- lished as analyses of gonnardite, no explanation was given as to why analyses originally published as analyses of a composite sample were republished 24 years later as analyses of a single mineral. In 1932 Hey had prorated the constituents found in these analyses be- tween a high-silica thomsonite and hypothetical gonnardite on the basis of assumed compositions and of the proportion of each estimated to be present in the spherule analyzed. The half-cell atomic ratios for this calculated gonnardite, D(IV), are given in table 3. Also given in table 3 with their atomic ratios, are analyses A, B, C, and D of Meixner, Hey, and Moss (1956), two other analyses of gonnardite from the literature, and three recent analyses of high-Na mesolite. The Bulgarian material (Kostov, 1958) analyzed was called “gonnardite (with thomsonite?).” The analysis of the material from Norway is an old one by Paijkull (1874). This material was originally called ranite, but Mason (1957) recently identified it as gonnardite. The analyses of high-Na mesolites given in the last three columns of table 3 indicate compositions very similar to that of gonnardite or high Na thomsonite. The relation between the atomic ratios of Ca and Na given in table 2 is shown graphically in figure 3. Also shown in figure 3 is the relation between Ca and N a in TABLE 3.—Analyses of gonnardites and high-Na mesolites, and their atomic ratios [In order of Increasing Nazo content] Gonnardites High-N a mesolltes D D (IV) C 1 2 A B 3 4 5 Composition (percent) 810- 41. 85 39. 21 42. 80 43. 20 40. 03 41. 15 40. 59 A110: 27. 02 31. 79 28.15 27. 90 27. 88 29. 49 29. 69 05:0 9. 29 5. 07 4. 26 3. 61 6.03 5.33 5. 06 NaaO ______ 7. 25 11. 55 12. 65 13.16 10.05 11.02 11. 00 K20 n.d. .......... .13 Trace .40 . 25 . 51 H10- __ 14. 37 11. 71 11.85 11. 74 1 11. 10 13. 52 13. 58 Total 99.78 ‘ 99. 90 99. 84 99. 61 5 100. 29 100.76 100. 43 Atoms Si ________________________________________________________________ 5. 64 5. 18 5. 62 5. 67 5. 48 5. 45 5. 41 A1 ________________________________________________________________ 4. 29 4. 95 4. 36 4. 32 4. 50 4. 60 4. 66 (Ca+Mg)- 1. 34 . 72 . 60 . 51 . 92 . 76 . 72 Na+K)..-. ....... .. 1.89 2. 96 3. 24 3. 35 2. 73 2. 86 2. 92 ECaO+N N) 3. 23 3. 68 3.84 3. 86 5 3. 65 3. 62 3. 64 determined). . 6. 46 5. 16 5. 19 5. 14 5.07 5. 97 6. 03 H20 Oécalculated) ._. _ 5. 91 5. 12 5. 04 4. 88 5. 49 5. 14 5. 08 Diflereuce... +. 55 +. 04 +. 15 +. 26 —. 42 +. 83 +. 95 1 Includes 0. 45 1130—. 1 Includes 3.12 11,0— ! Includes 0.18 MgO (0.04 Mg atomic ratio). 4 Includes 0.57 FeaOa. IIneludes 0.85 Feaoa, 0.59 FeO, 0.22 M30 (0.04 Mg atomic ratio) and 0.02 T103. LOCALITY AND ZREFERENCE FOR ANALYSES IN TABLE 7 D. Chaux de Bergonne, France, Meixner, Hay, and Moss, 1956,}; D(IV). Hypothetical composition of gonnardite, calculated by Hey (1932, p. 117) from D. Aci Casteilo, Sicily, Meixner, Hey, and Mouse, 1956, p. 266. 1.Bonrgass , aria, Kostov, 1958, p. 16, 2Langesun flord, ven, Norway, Dana, 1892, p. 609 (ranite). A. Klock, Sty-tie, Meixner, Hey, and Moss, 1956,p .266. B. Aci Trezza, Sicily, Meixner Hey, and Moss, 1956, p. 266. 3. Rio Cambone, Montifen‘o, éardinia, Deriu, 1954, p. 42. 4, 5. Kladno, Bohemia, Antonin, 1942, p.1 Na ATOMIC RATIO COMPOSITIONAL RELATIONS AMONG THOMSONITES, GONNARDITES, AND NATROLITES 4.0' k‘l’heoretical \ natrolite \ \ \ O \ \ 3.6 Q \ \ EXPLANATION Upper limit of \ / gonna‘rdite \ x ’ x3 ‘ \ Gonnardites /' A \ \ . 3.2 X xNorway\\¢ High—Na thomsonites '23 \3\ a \ O High—Na mesolites \ \9‘ “k 0. \ \so Natrolltes 2 s ‘3} ' ’1’ o \9®_ . 5 \7/ Q09 \ ‘ «7,, 0% \ \°<% I‘m \ o o \ 2 4 \9’58 60 \ ’ \91 x "5 ‘ e, \ , \ \ \ 66. \ XC \ \ \ 2.0 9K \ ‘ He '5 onnardite y 8 \ Dx \ Winchell's d me b ‘ ' en m er A\ 6. .9 \Lower limit of / gonnardite 1 6 ‘ 69. \ . 24a \ Theoretical faroelite * \ Hey's high-Si thomsonitefl~ Average faroelite }9\ \ \ 1-2 \ \ \ Theoretical /thomsonite \ \\ 0.8 ‘ " \ \ \ \ \ \ \ 0.4 \ \ \ \ \ \ \ \ 0 0.4 0.8 1.2 1.6 2.0 2.4 Ca ATOMIC RATIO FIGURE 3.——Relation between Ca and Na. in gonnardites, and high-Na. thomsonites and mesolites. E6 the high—N a thomsonites given in tables 5 and 6, Nos. 1, 5, 6, 9, 24a, and 66. The relation between Ca and Na in No. 28, which had the highest Na content of the thomsonites cited by Hey (1932), is also shown in figure 4. The material represented by this analysis had previously been called mesolite. The points represent- ing gonnardites intermingle with points representing high-N a thomsonites and mesolites, and all the points fall close to and along the line representing dual replace- ment, in equal amounts, of Ca by N812 and Ca as CaAl by NaSi. On the other hand, the point representing the relation between Ca and Na in the gonnardite D(IV), calculated by Hey frOm analysis D falls on the line representing NaSi—eCaAl replacement at a point equivalent to 50 percent replacement of the Ca in thomsonite as CaAl, and a step beyond the point repre- senting theoretical faroelite, which is equivalent to a replacement of 25 percent of the Ca in thomsonite as CaAl. The point representing the thomsonite calcu- lated by Hey, D(III), also falls on the NaSiaCaAl replacement line a little below the point representing theoretical faroelite. Thus the points for “gon- nardite” and “thomsonite” calculated by Hey from analysis D both fall on the N aSiaCaAl replacement line, whereas the point representing D itself falls much closer to the line representing dual replacement. The intermingling of points representing the relation between Ca and Na in gonnardites, high-N a thomson- ites, and so—called mesolites along the line representing dual replacement of Ca in thomsonite suggest that all these zeolites may belong to an isomorphous series characterized by dual NaSi—>CaAl and Na2eCa replacement, and that theoretical thomsonite, CaHNal,oAl5loSi5,OOZO-6.0 H20, is the high Ca end . member of this series. A THOMSONITE-GONNARDITE-NATROLITE ISOMORPHOUS SERIES? Gonnardites and high-N a thomsonites can both be interpreted as derived from normal thomsonsite by dual replacement, to about the same degree of Ca by Nag, and of CaAl by N aSi. Steps in this dual replacement, starting with thomsonite, are shown in these formulas. Thomsonite: Nal_oCa2.oA15.oSi5.0020- 6.0 H20 Na1,75 031.50A14.7ssi.5.2502o ' 5-5? H20 Nazso 031.00Al4.5OSi5_5002o - 5.0? H30 N33.25C30.50A14.25 Si5.75020'4-5? H20 Natrolite: NameCaomAlmoSimoOzo - 4.0 H20 The atomic ratios for No. 9, table 1, are very similar to those in the second formula, which represents the first step in the replacement. The atomic ratios for SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY the analyses of gonnardite from Bulgaria and the high- N a mesolite from Sardinia, table 3, are similar to those in the third formula, and the atomic ratios for analysis A and B, table 3, are similar to those in the fourth formula. The Ca-free end member is identical with the formula for natrolite. Points representing the relation between Ca and Na in several natrolite analyses, Nos. 12, 15, 17, 22, and 25 (table 1), fall on or close to the upper end of the line representing dual replacement in figure 3. The gonnardite (B, table 3) having the highest Na content has a Na atomic ratio only 0.26 below the Na atomic ratio in the natrolite having the highest Ca content CNo. 25, table 1). for these two analyses are: Analysis B, table 32 Nas,3aCa,51Al4_3gsi§,o7020' 5.14 H30 Analysis 25, table 1: Na3.alCa,z7AI4.zgSi5_77Ozo- 4.37 H20 The sum of the cations (Ca+Na) also agreesclosely, 3.86 and 3.88 in B and N0. 25, respectively. The high—Na mesolites given in table 3 may be interpreted as either derived from thomsonite by replacement of 0.40—0.50 CaAl by 0.40—0.50 NaSi, plus replacement of 0.60—0.65 Ca by 1.20—1.30 Nag, or as derived from mesolite by replacement of about 0.50 CaSi by an equivalent amount of NaaAl. Deer, Howie, and Zussman (1963, p. 358—359) give the following ranges in indices of refraction for thomson- ite, gonnardite, natrolite, and mesolite: Optic axial a 3 7 sign Thomsonite ______________________ + 1. 497—1. 530 1. 513—1. 533 1. 518-1. 544 Gonnardite. _ _ _ _______ — 1. 497—1. 506 ______________ 1. 499-1. 508 Natrolite ______________ + 1. 473—1. 483 l. 4764. 486 1. 485-1. 496 Mesolite _________________________ + .............. 1. 504—1. 508 ____________ The following data are available on some high—N a thomsonites whose analyses are given in tables 1 and 2 and on the gonnardites and the high-N a mesolites whose analyses are given in table 3: Analysis at B -y Source High-Na thomsonites 1.513 1.518 Hey, 1932, p. 54. 1. 525 1. 527 Do. 1. 520 1. 528 Do. 1. 519 1.526 Do. 1. 523 1. 528 Hey, 1932, p. 80. Gonnardites and high-Na mesolites Meilxper, Hey, and Moss, 1956, p. 266. o. B 1.... . .. _._ . C 1 _______________ . _. ___ . Do. D Do. Norway 1 ........ 1 Mason, 1956. Sardinia _________ Deriu, 1954. I The indices of refraction given are not true a and true 7, but a' and 1'. The formulas _ , COMPOSITIONAL RELATIONS AMONG THOMSONITES, GONNARDITES, AND NATROLITES 1.536 1.532 1.528 1.524 1.520 ‘MEAN INDEX OF REFRACTION 1.516 1.512 18 .0 18.4 18.8 19.2 19.6 AI ATOMS PER UNIT CELL 20.0 20.4 FIGURE 4. Relation between A1 atoms per unit cell and mean index of refraction in thomsonites. Hey (1932) found that the indices of refraction of thomsonite increase with increase in Al, but are little affected by replacement of Ca by N32, or vice versa. The relation between A1 atoms and Na atoms per unit cell and the mean index of refraction in 16 thomsonites whose analyses were reported by Hey (1932) are shown in figures 4 and 5. The general trend of the points in figure 4 is upward with increase in Al content and mean index of refraction, but the relation is not precise, as some thomsonites having the same Al content may differ considerably in mean index of refraction and vice versa. The location of the points in figure 5 indicates little relation between Na content and mean index of refraction. For example, six thomsonites having about the same number of Na atoms per unit cell, 4.74 to 5.04, range in mean index of refraction from 1.521 to 1.535. Thus the wide ranges in the indices of refraction of thomsonite seem to reflect the wide range in Al content, or conversely, Si content, in thomsonites, rather than variations in Na content or Ca content. Gonnardites are characterized by varying NaSi—e CaAl and Nag—>Ca replacement, and should vary con— siderably in Al content, and, presuming the same relation as in thomsonite, in indices of refraction. The ranges in the indices of refraction for gonnardite given by Deer, Howie, and Zussman are coextensive with the ranges for prime (’) indices of refraction of gonnardite reported by Meixner, Hey, and Moss (1956), and are presumably based on them. 'Thus the ranges given by Deer, Howie, and Zussman for the indices of E7 1.536 I O O 1.532 I Z « O o - ; 1.528 2 O 0 1: LL LLJ Di '5 1.524 5 ' - g o o . _ I Z 5 1 520 2 1.515 U . 2 1 512.0 4.0 6.0 8.0 10.0 12.0 Na ATOMS PER UNIT CELL FIGURE 5. Relation between Na atoms per unit cell and mean index of refraction in thomsonites. refraction of gonnardite are based on four sets of oz’ and 7’ values. Although these four gonnardites vary widely in Na content, from 1.89 to 3.35 atoms per half cell, they are very similar in Al content, varying only between 4.29 and 4.36 atoms. This is because in D and C, which have the lower Na contents, 1.89 and 2.12 atoms per half cell, replacement of CaAl by NaSi is greater than replacement of Ca by Naz, whereas in A and B, which have the higher Na contents, 3.24 and 3.35 atoms per half cell, replacement of Ca by Na2 is greater than replacement of CaAl by N aSi. In all four the amount of NaSi replacement of CaAl is about the same, 0.62+0.7O CaAl replaced by equivalent amounts of N aSi, but the amount of Ca replaced by Naz varies from 0.23 in D and C, to 0.84 in A and 0.86 in B. Consequently all four have about the same Al content and about the same indices of refraction. The relative degree of NaSi or Naz replacement is indicated by where the points representing these gon- nardites fall with respect to the dual replacement line in figure 3. The prime (’) indices of refraction reported by Mason (1957) for the Norwegian zeolite that he identified as gonnardite are somewhat higher than those reported for analyses A, B, C, and D, as its higher Al content would suggest, 4.95 atoms per half cell. The indices of refraction reported by Hey for the high-Na thom- sonites whose points intermingle with those of gon- E8 nardite along the dual replacement line in figure 3 are also higher than those for gonnardites A, B, C, and D. The high-Na thomsonites are also higher in Al content than these gonnardites, having relatively less replacement of CaAl by N aSi—from 0.24 to 0.46, compared to 0.62 to 0.70 in A, B, C, and D. The indices of refraction for these high—Na thomsonites fall within the ranges of indices for thomsonites. It is probable that further data on the indices of refraction of gonnardite will extend the ranges of the indices of refraction of this zeolite and may show that the ranges of the indices of refraction for gonnardite overlap to some extent those for thomsonite. The ranges for the indices of refraction of natrolite and mesolite are very narrow, just as their variations in Al content are very small. The structural similarities between thomsonite, gonnardite, and natrolite are well known. All three zeolites have framework structures in which every (Si, ADO; tetrahedron has each of its oxygens shared with another tetrahedron; linkages to form chains in the z direction are the most prominent. The repeat unit in each chain consists of five tetrahedra and occu- pies about 6.6A. However, natrolite and thomsonite differ considerably in detail. The structures of thomso- nite and natrolite are described by Deer, Howie, and Zussman (1963, p. 361—362) “In thomsonite neighbor- ing chains in the y direction are related by mirror planes (010), but those in the a; direction are related by diad axes [010]. The resulting unit cell (orthorhombic ana) contains four chains * * * . In the structure of natrolite, neighboring chains are related (approxi- mately) by diads only, and this results in a body— centered cell with dimensions similar to those of thomsonite (azbz 13.1.4) or a face-centered cell with a. 18.3, b 18.63 * * * . The difference between A1 and Si tetrahedra and perhaps other deviations from the ideal structure result in orthorhombic (pseudo— tetragonal) cells for both thomsonite and natrolite. In thomsonite, moreover, such deviations give rise also to a double 6 parameter (13.25z2X6.63).” The structure of gonnardite is described by Deer, Howie, and Zussman (1963, p. 364) as follows, “Gonnardite has cell parameters similar to those of thomsonite and is assumed to have a similar aluminosilicate framework structure. X-ray fibre photographs of gonnardite are similar to those of thomsonite, and its powder photo- graphs may be confused with those of natrolite (Meixner et al., 1956).” N atrolite is the only one of these three zeolites whose structure has been refined (Meier, 1960). As the structures of these minerals are similar but differ in detail, it may be that, although analyses sug- gest a continuous series compositionally, there is a structural change somewhere between thomsonite and SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY natrolite. Such a change would be analogous to the structural change from dioctahedral to trioctahedral which occurs halfway between muscovite and lepidolite, and which breaks the continuity of the compositional series consisting of muscovite, lithian muscovite, and lepidolite (Foster, 1960). Hey (1932) concluded that the optical properties and X—ray spacings of gonnardite furnish fairly conclusive evidence that it should be regarded as a separate species related to thomsonite perhaps in a manner like that of a— and B—quartz. He tentatively identified gonnardite with metathomsonite, a high-temperature polymorph of thomsonite. The transition from thomsonite to metathomsonite is reversible and the transition temper- ature is dependent on the water content. The two forms are compositionally the same, except for water content. On the other hand, gonnardite differs in composition from thomsonite, being intermediate in composition between thomsonite and natrolite, and cannot become thomsonite simply by a decrease in temperature and rehydration as metathomsonite does. To obtain thomsonite from gonnardite requires chemical change. Hey (1955, p. 166) gives the H30 content of gon- nardite as 7 molecules per 20 oxygen atoms. The next year Meixner, Hey, and Moss (1956) give the H20 content as 6 molecules per 20 oxygen atoms. It has been noted that in natrolite and mesolite the ratio of H20 molecules to Na ions is 1:1, and that in scolecite and mesolite the ratio of H20 molecules to Ca ions is 3:1. In thomsonite the ratio of H20 molecules to Na ions is again 1:1, but the average ratio of H20 molecules to Ca ions is 2.5: 1. Analogously, it would be expected that gonnardites would contain less water than thom- sonites, because of the higher Na content and lower Ca content. Also, as they are intermediate in chemical composition between thomsonite and natrolite, they would also be expected to be intermediate in H20 content. In the analyses of gonnardite given by Meixner, Hey, and Moss, the water content per 20 oxygen atoms varies from 5.35 to 6.1, molecules. In the analyses given in table 3, which includes four of the analyses of Meixner, Hey, and Moss, H20 varies from 4.85 to 6.46, and averages 5.54 molecules. The average molecular ratio for H0 in the high-N a thomsonites that fall close to or on the middle line (Nos. 5, 9, 248., 66, and 69) is 5.42. The average molecular ratio for H20 in all the analyses whose Ca:N a ratio falls on or close to the dual replacement line in figure 3 is 5.50. For 10 of these 15 analyses the molecular ratio for H20 is less than 5.50, in three it is close to 6.00, and in two it is more than 6.40. Such H20 values tend to support the hypothesis that in zeolites that are intermediate in composition between thomsonite and natrolite the COMPOSITIONAL RELATIONS ’AMONG THOMSONITES, GONNARDITES, AND NATROLI’I‘ES content of H20 should also be intermediate, and that H20 contents of 6.00 or 7.00, as assumed in formulas for gonnardites, are too high. In this connection it should be taken into consideration that the average given above is based on total H20, as only two of the analyses upon which this average is based reported H20—. Therefore the true H30 content of these zeolites is probably somewhat lower than this average. FORMULATION The formula given by Hey (1955) for gonnardite, 2 [NazCaALSieOzoJHaO], can be derived from thom- sonite only by unilateral replacement of CaAl by N aSi, as illustrated below: Thomsonite: Nal,oCaz.oA15_oSi,5_oOgo-6.0 H20 Faroelite: N31 .5 Ca] .5A14,5Si5.5020-6.0 H50 Hey’s gonnardite: Nag,oCa1 ,oA14.oSio,0030-7.0 H20 However, the atomic ratios of all the analyzed gon- nardites, as well as the atomic ratios of most of the high Na—thomsonites, indicate dual replacement; re- placement of CaAl by NaSi, and replacement of Ca by N a; to about the same degree. Therefore, the formula given by Hey for gonnardite does not reflect the com- position of gonnardite as indicated by analyses, nor does it reflect the great variation in composition found in the materials that have been called gonnardites. Nor does the formula usually given for thomsonite re- flect the range in composition found in the thomsonites. The usual thomsonite formula is that of the end mem- ber, which bears the same relation to thomsonites in general as phlogopite bears to phlogopites in general, and which bears the same relation to gonnardite that phlogopite bears to biotites. It is obvious that members of a replacement series that vary in composition and that merge into one an- other cannot be adequately characterized by formulas that indicate fixed relations between the constituents. For members of such a series only formulas that indi- cate permissible limits of composition are adequate. The limits of composition with respect to Na and Ca between thomsonite and gonnardite and between gon- nardite and natrolite suggested herein are selected arbitrarily because the analyses indicate considerable continuity, with no natural breaks along the line joining thomsonite and natrolite (fig. 3). The suggested upper limit for thomsonite is at a dual replacement of 0.20 Ca+0.20 Al by 0.20 Na+0.20 Si and of 0.20 Ca by 0.40 Na. Starting with theoretical thomsonite, Na and Ca at this degree of dual replacement are both 1.60. Up to this degree of dual replacement Ca is greater than Na, beyond it Na is greater than Ca. At this degree of replacement of CaAl by NaSi the A] content is 4.80 and the Si content is 5.20. The upper limit of composition E9 for thomsonite along the line of dual replacement would be, therefore,Nal 500811. soA-l4 .soSia .20020 ' 5 ~60H20- This com- position also represents the lower limit of gonnardite. The suggested upper limit of gonnardite is at a dual re- placement of 0.80 Ca+0.80 Al by 0.80 Na+0.80 Si and of 0.80 Ca by 1.60 Na. This degree of replacement of CaAl by NaSi results in an Al value of 4.20 and a Si value of 5.80. The upper limit of composition for gon- nardite would be NagAoCaAoAh,zoSi5.80020-4.40. H20. The complete formulas with these suggested limits of composition for thomsonite, gonnardite, and natrolite would then be Thomsonite: Na, .0-1 .6033 .0-1 .aA15 ,0-4 .3815 .0—5 .20020-6.0—5.6 H20 Gonnardite: N81.t-a.4cai.t-o.4A14.s—4.2Si5.2—5.8020-5-6—4-4 H20 Natrolite: Naa .4-4 .0 080 .4—0 .0A14.2-4.OSi5.8-6.0020‘4-4'_4- 0 H20 These limits for thomsonite, gonnardite, and natrolite are marked in figure 3. In these formulas the values for H20 are calculated on the basis of a 1 :1 ratio of H20 to Na and of 2.5 : 1 ratio of H20 to Ca, as in thomsonite. In accordance with these range formulas, analysis No. 24a, table 2, with 1.58 Na and 1.59 Ca comes just Within the upper limits for thomsonite, and in 69, table 2, Ca is just within the lower limits for gonnardite. N0. 1, table 1, falls well within the limits for thomsonite. No. 9, table 1, and Nos. 28 and 66, table 2, and all the analyses given in table 3 fall within the gonnardite range. It must be understood that the range formula given above for thomsonite applies only to thomsonites belonging to this dual replacement series. Thomsonite in the NaSi—meCaAl line of replacement would have a range formula reflecting the atomic relations in this line of replacement, such as Cfi’2.0-1.5Na’l .0—1 .5A15.o—4.5Si5.0~5.5020'6-0—5-25 H20: which embraces faroelite. OTHER LINES OF REPLACEMENT In addition to the two lines of replacement that have been discussed, four other possible lines of replacement are indicated in figure 3. One of these, like the two that have been discussed, is a Ca replacement line, replacement of Ca by N82. The other three are all Na replacement lines, continuations of the three Ca replacement lines. However, except for a few points that fall close to theoretical thomsonite, none of the points representing the Ca/Na relations of analyses whose atomic ratios are given in tables 1 and 2 fall along these lines. Replacement of Na by Ca in thomsonites appears to be quite subordinate to replace- ment of Ca by Na, and replacement of Ca by N a2 seems to be found most commonly in conjunction with replacement of Ca as CaAl by N aSi, as along the dual replacement line. i E10 CONCLUSION This study has shown that gonnardites, and several high-Na mesolites, and certain thomsonites that are characterized by dual NaSi——>CaAl and N ar—éCa replacements, are all intermediate in composition be- tween thomsonite and natrolite, and suggest an isomor- phous series between them. However, differences in optical properties and in structure cast doubt on such an interpretation. Whether all these zeolites are members of an isomorphous series between thomsonite and natrolite or whether only the high-Na thomsonites belong to such a series, and the gonnardites and high-Na mesolites are polymorphic forms, are problems that require more data, especially on their structures, and more study for their clarification. REFERENCES Antonin, Rudolf, 1942, Research on the minerals and rocks of Vinaricka hill: Kralovské Ceské Spoleénosti N auk Véstnik, art. 2, 19 p. Deer, W. A., Howie, R. A., and Zussman, Jack, 1963, Rock forming minerals. v. 4, Framework silicates: New York, John Wiley & Sons, Inc., 435 p. Deriu, Michele, 1954, Mesolite di Rio Cambone (Montiferro- Sardegna centro-occidente): Periodico Mineralog. Roma, v. 23, p. 37—47. 0 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY Foster, Margaret D., 1960, Interpretation of the composition of lithium micas: U.S. Geol. Survey Prof. Paper 354—E, p. 122—127. ' ‘ . Hey, Max H., 1932, Studies on the zeolites. Part II. Thom- sonite (including faroelite) and gonnardite: Mineralog. Mag. v. 23, p. 51—125. ————— 1955, An index of mineral species and varieties arranged chemically, 2d ed.: London, British Museum (Natural History), 728 p. Kostov, Ivan, 1958, Zeolites in Bulgaria: scolecite, mesolite, “gonnardite,” and thomsonite: Sofia Univ. Biolog—Geolog- Geograph. Fakultet, Godishnik, v. 53, p. 1—24. Mason, Brian, 1957, Gonnardite (ranite) from Langesundsfjord: Norsk Geol. Tidsskr., v. 37, p. 435—437. Meier, W. M., 1960, The crystal structure of natrolite: Zeitschr. Kristallographie, v. 113, p. 430-444. Meixner, Heinz, Hey, M. H., and Moss, A. A., 1956, Some new occurrences of gonnardite: Mineralog. Mag. v. 31, p. 265— 271. Paykull [Paijkull], S. R., 1874, Rauit, ein neues Mineral von Brewig: Deutsch. Chem. Gesell. Berlin, Ber., v. 7, pt. 2, p. 1334—1335. Taylor, W. H., Meek, C. A., and Jackson, W. W., 1933, The structure of the fibrous zeolites: Zeitschr. Kristallographie, v. 84, p. 373-398. Winchell, A. N., 1925, A new theory of the composition of the zeolites: Am. Mineralogist, v. 10, 88-97. ———~— 1926, Doubtful mineral species as illustrated by “faroelite”: Am. Mineralogist, v. 11, p. 82—89. 0/5 75" 7 DAY 1% VI 5'0 “/ ‘F , . Underground Temperatures ,_ and Heat Flow in the East Tintic District, Utah GEOLOGICAL SURVEY PROFESSIONAL PAPER 504—F OFAEOAL ~ Q5“ Y A‘ A \‘R’ 0“,; Q § MAY 41956 <2 S“ ’97” scrmct “‘5“ “ Underground Temperatures and Heat F 10W in the East Tintie District, Utah By T. S. LOVERING and H. T. MORRIS SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY GEOLOGICAL SURVEY PROFESSIONAL PAPER 504—F Tecam'gaes ana’resa/tr qf a geotaerma/ staa’y of an area aa‘vz'ag saértaatz’a/ temperature anal amt-flaw anomalies related to saéterraaeaa fiat springs and geologic rtractare UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1965 UNITED STATES DEPARTMENT OF THE INTERIOR STEWART L. UDALL, Secretary GEOLOGICAL SURVEY Thomas B. Nolan, Director For sale by the Superintendent of Documents, U.S. Government Printing Oflice Washington, DC. 20402 CONTENTS Page Abstract ___________________________________________ F1 Conductivity of rocks—Continued Page Introduction _______________________________________ 1 Conductivity of rocks and formations in the East Measurement of temperatures ________________________ 2 Tintic district ________________________________ F10 Surface temperatures ____________________________ 2 Calculation of temperatures at depth __________________ 11 Temperature gradients in deep holes ______________ 2 Geologic factors ________________________________ 11 Temperatures in mine workings ___________________ 5 Calculation of temperatures at the water table _____ 12 Temperature gradients, conductivity, and heat flow _____ 6 Temgerature 0f ground water ------------------------ :3 Relation of thermal conductivity, temperature eneral features """""""""""""""""" 3 . East Tintic thermal area _________________________ 14 gradient, and heat flow ________________________ 6 Heat loss 14 Heat flow and thermal gradients in uniform sossk Scum... .;as:::::::::::::::::::::::::::::::: 14 constant nonuniform temperature at depth ------- 7 Approximate flow of subterranean hot springs ______ 18 Conduct1v1ty 0f 1‘ ocks ------------------------------- 8 Origin of thermal waters _________________________ 20 Direct determination of conductivity ______________ 8 Latite Ridge thermal area _______________________ 26 Indirect determination of conductivity ____________ 8 References cited ____________________________________ 27 ILLUSTRATIONS [Plates are in pocket] PLATE 1. Map of East Tintic thermal area, Utah, showing isotherms representing mean annual temperature at the surface, major geologic features at the water table, and surface and subsurface thermal data. 2. Chart showing lithology, temperatures, and gradients in drill hole GS 1 in East Tintic district, Utah, at various times, and conductivities calculated from gradients. 3. Map of East Tintic thermal area, Utah, showing isotherms and major geologic features at the water table, and surface and subsurface thermal data. 4. Map of the Eureka quadrangle showing generalized surface geology, elevation of ground-water surface, and area assumed available for recharge of ground water in East Tintic thermal area, Utah. 5. Map of East Tintic thermal area, Utah, showing heat loss at the surface, major geologic features at the water table, and surface and subsurface thermal data. Page FIGURE 1. Graph of isotherms in homogeneous rock layers above temperature transition zone ___________________________ F9 2. Analog-model—simulator graph showing distribution of isotherms in homogeneous rock above hot and cool zones- _ _ _ 9 3. Graph of ratios of SiOz to total solids, Br:Cl, and K: Na in subsurface water ________________________________ 22 TABLES Page TABLE 1. Thermal conductivity of rocks and formations of the East Tintic district as determined by various methods _____ F10 2. Excess heat flow in East Tintic area north of Apex Standard fault zone ____________________________________ 15 3. Chemical analyses of water from Eureka city well, Burgin mine, North Lily mine, and of saline hot springs of Utah having temperatures above 115°F ___________________________________________________________________ 16 4. Spectrographic analyses of water samples from Burgin mine and Abraham Hot Springs _______________________ 18 5. Variation in chloride content, temperature, and relative deuterium content of water samples from Burgin mine__ __ 19 6. Precipitation and ground-water recharge in the East Tintic district ________________________________________ 20 7. Chlorine and bromine data for evaporating sea water and the resulting solid phases __________________________ 21 8. Lithium to sodium ratios of various chloride-type waters _________________________________________________ 25 III SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY UNDERGROUND TEMPERATURES AND HEAT FLOW IN THE EAST TINTIC DISTRICT, UTAH BY T. S. LOVERING and H. T. Monms ABSTRACT The deep water table in the East Tintic district slopes very gently to the east at approximately 4,550 feet above mean sea level and is from 1,000 to 1,500 feet below the surface. The temperature of the water at this elevation differs greatly from place to place, and measured temperatures range from approx— imately 80°F in the western and northwestern parts of the district to 140°F in the south—central part. Rock temperatures were measured in drill holes, ranging from 33 to 2,400 feet deep, and at many places in the extensive mine workings present in the central and southern parts of the district. Thermal conductivities were determined in the laboratory On many individual specimens, and on formations in the field by various methods. From these data, temperature gradients, heat loss, and mean annual temperatures at the surface were calculated. These data also enabled us to calculate the temperature at the level of ground water even where points at which temperatures were measured are well above or below water level, provided that the conductivities of the rocks between the point of temperature measurement and water level were known. Temperatures either measured or calculated were plotted for the water table over a wide area and a sufficient number of such points are available to show the general thermal pattern on maps. The theoretical basis for these calculations and the techniques used for measuring temper- atures are described, and the results are shown on maps by means of isotherms representing mean annual temperatures at the surface, temperatures at water level, and isograms showing anomalous heat flow in microcalories per square centimeter per second in the East Tintic thermal area. The geographic distribution and geologic relations of the thermal waters indicate that a steep northward-trending fracture zone in the footwall of the East Tintic thrust fault acts as a conduit for rising hot water which in part follows the overlying thrust fault and in part moves upward through short-circuiting fissures in the hanging wall. This water rises through much- fractured rock that is saturated with ground water, which in general is moving toward the east or northeast. The subter- ranean hot springs, however, apparently cause local reversals in the direction of ground-water flow. Although the highest temperature measured at water level is 140°F, the calculated temperature at water level below the southernmost drill hole entered (EP 2) is 163°F. This hole is in the Latite Ridge thermal area, an area that to date (1964) has been inadequately explored, and therefore most of the quanti- tative estimates of temperature and heat flow are given for the main East Tintic thermal area which lies north of the Apex Standard thermal trough. However, the temperature rises southeastward at a uniform rate of 2.5°F per 100 feet horizontally through a distance of 2,000 feet (to ET 2) and if this rate held for another quarter of a mile the ground water there would be at the boiling point. In the main East Tintic thermal area, the average temperature of the anomalously hot water is about 104°F, and an estimate of the approximate annual recharge of water at a temperature of 80°F allows calculation of the amount of water, 100 gpm (gallons per minute), at 143°F, which is the highest temperature calculated for ground-water level, required to raise its heat content to that of water at the average temperature of 104°F. The heat loss from the rock cover can also be calculated and is equivalent in heat energy to about 120 gpm of water at 143°F. The sum of these two waterflow rates suggests that subterranean hot springs in the East Tintic district contributes about 200 gpm of water if the temperature of the hot water in the conduit underlying the East Tintic thrust fault approximates 143°F. The hottest waters sampled are saline and apparently belong to a family of saline hot springs that are found at several places in Utah. The ratios of SiOzzsolids, Br: Cl, KzNa, and LizNa all suggest a magmatic contribution, and the deuterium content of the water clearly precludes its being hot connate water. The evidence now available indicates either that the heat and mineral content of the hot saline water is of deep volcanic origin or that meteoric water moving through Jurassic salt beds below two or more major thrust faults is mixed with volcanic ema- nations at depth. The high gradients in the Latite Ridge thermal area, together with a seemingly favorable geological structure, suggest that the area should be explored as a possible source of geothermal power, but the thermal data now available (1964) are inadequate for an appraisal of this possibility. INTRODUCTION The East Tintic district, which is about 60 miles south of Salt Lake City, Utah, is well known to mining men not only for the rich blind ore bodies that have been exploited there but also for the great range in temperatures in the mine workings, most of which are above the permanent ground-water level. The ground- water level, which is at an elevation of approximately 4,550 feet, is from 1,000 to 1,500 feet below the surface throughout much of the district, and the mine workings above the water level are dominantly dry and well suited for measuring temperatures that are undisturbed by moving ground water except quite locally. During F1 F2 the period 1943-63 we were able to measure under- ground and surface temperatures at many places while studying the geology of the district. The information on temperatures gathered during this time is summa- rized on the maps that accompany this paper and is the basis for the conclusions reached concerning heat flow, heat sources, and the underground temperature distribution in the East Tintic district. The study of the surface temperatures, which was an essential preliminary to the present report, has been described in another paper (Lovering and Goode, 1963). Drill-hole and other underground temperatures were recorded by many members of the field parties who participated in the geological study of the district from 1943 to 1950, but we assume the responsibility for the accuracy of all measurements reported here and regret that individual credit cannot be given for each measure- ment used. However, we wish to thank Robert H. Roy for the detailed temperature logs of holes BC 5 and BC 90 (see also Roy 1) and the Bear Creek Mining Co. for temperatures from several of their deep drill holes. MEASUREMENT OF TEMPERATURES SURFACE TEMPERATURES The method of ascertaining mean annual surface temperatures, which has been described in detail (Lovering and Goode, 1963), depends primarily on extrapolating to the surface the temperature gradient determined in shallow or deep holes. Direct measure— ments of temperature gradients at depths of less than 100 feet are in general unreliable unless the temperatures measured are recalculated to include the effect of the annual temperature wave at the surface. At depths greater than 100 feet, two or more temperature meas- urements should indicate the rate of change of tempera- ture with depth; if the temperature gradient so ascertained is then extrapolated to the surface, the resulting figures should approximate closely the mean annual temperature at the surface unless the conduc- tivity of the rocks changes drastically through a sub- stantial part of the section above the points where the gradient was measured. In few places in the East Tintic district does the lower. conductivity of the weathered layer near the surface suffice to change by as much as 1°F. the temperature extrapolated from below a depth of 100 feet. Surface temperatures are determined almost entirely by climatic, geographic, and vegetational features and show virtually no relation to deep underground heat sources. The exposure of slopes to sun and wind are major factors governing the mean annual tempera— 1 Roy, R. F., 1963, Heat flow measurements in the United States: Harvard Univ. unpub. Ph.D. thesis. SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY ture, but the vegetational cover and the snow cover are also of importance. A dry bare, south-facing slope may have a mean annual temperature several degrees Fahrenheit higher than that of an adjacent north- facing slope covered with pinyon pine or brush. The surface temperatures as determined from many shallow and deep drill holes, together with the inferred iso- therms representing mean annual temperatures at the surface, are shown on plate 1. TEMPERATURE GRADIENTS IN DEEP HOLES The measurement of temperatures in deep drill holes appears deceptively simple. The most common major error in such measurements is caused by the assump- tion that if the probe has been held at a point long enough to come to the temperature of the water or air surrounding it, the temperature measured by the probe is the true rock temperature. A thermometer in a cold metal case that occupies a substantial part of the drill hole at the point of measurement may greatly lower the temperature of the air or water surrounding it. A difference of only a few degrees Fahrenheit between the case and the surroundings has little effect on the recorded measurement, whereas a difference of several tens of degrees, such as may exist between the temperature of the equipment at the surface on a cool day and the temperature measured underground, would materially lower the temperature reached by the maximum thermometer. A probe lowered rapidly on a cold February morning to the bottom of one deep drill hole, some 625 feet below ground-water level, and left there for 45 minutes gave a reading that was 20°F too low. Unless the probe can be left in place for several days, in-hole measure- ment at given depths should always be a two-step procedure; the probe should first be held a short dis- tance—15 to 30 feet—~above the point of measurement long enough for the probe to come reasonably close to temperature equilibrium with the surrounding air or water, and then should be lowered to the point of temperature measurement. In water, the time required to reach equilibrium commonly ranges from 3 to 15 minutes, whereas in air the time required is generally at least 1 or 2 hours, and a halt of several hours is desirable. The actual time required depends on the accuracy sought, on the kind of probe employed, and on the material used to protect or enclose the probe. The requisite time for satisfactory results with a give a given type of probe should be determined by trial in the field. The probable error inherent in measuring a high tem- perature underground with cold equipment has just been noted; the converse is also true. If drill-hole tem- peratures are being measured on a warm summer day, UNDERGROUND TEMPERATURES AND HEAT FLOW, EAST TINTIC DISTRICT, UTAH F3 the equipment must be prechilled, and a maximum thermometer must start its downward journey showing a temperature well below that which is measured under- ground. We used a Dewar flask filled with ice and water to provide cold water for chilling the rag-wrapped thermometer and then whirled the thermometer at the end of a 3-foot string to lower the mercury column to the desired level. The thermometer was then quickly placed in a chilled case and lowered rapidly for the first 50 feet below the surface. Each time the thermom- eter was brought to the surface to be read, the thermom- eter case was quickly immersed in cold water and opened, and the thermometer itself was withdrawn only far enough to read—before the summer heat could affect it. Where an accuracy of i0.1°F is sufficient, Well- calibrated maximum mercury thermometers in water- tight probes are quite satisfactory, though somewhat difficult to use in hot weather. Most of the drill-hole temperatures in the East Tintic district were first measured with maximum ther- mometers calibrated by the Bureau of Standards and believed accurate to i0.1°F. The equipment used was that devised by C. E. Van Orstrand while he was Chief Physicist of the US. Geological Survey (Van Orstrand, 1924). A bronze tube about a foot long, with watertight screw cap, held one or more thermom- eters and was attached to piano wire that ran over a grooved calibrated wheel meshed with a counter that showed the depth of the thermometer in feet. A small sash weight was attached to the bottom of the brass tube to provide tension on the line at all times and to prevent the wire from snarling when lowered into an an open hole with irregular sides. This weight also signaled the presence of obstructions or the bottom of a hole by the sudden drop in tension on the line. For accuracy greater than 0.1°F, perhaps the two best devices are (1) the bathythermograph and (2) the resistance thermometer (of the type that utilizes a small thermistor designed for the range of temperatures expected). The bathythermograph, being of larger diameter than either the maximum or resistance ther- mometer, has the disadvantage of requiring a hole at least 4 inches in diameter; it is most useful where air temperatures are to be measured and no pressure cor- rections are required. If the bathythermograph is well calibrated, it is a very satisfactory instrument but re- quires more time to reach equilibrium than do either the maximum thermometor or the resistance thermom- eter. We used a bathythermograph somewhat similar to those designed by A. F. Spilhaus for oceanographic work (Spilhaus, 1938). In many ways a resistance thermometer having a well-calibrated thermistor is the most satisfactory probe for measuring temperatures in a drill hole. In deep substantial errors in the measurement. drill holes, however, the weight of the electric cable causes a considerable stretch in a new line, and the length of wire going into a deep hole should be checked against the length that comes out, so that a correction factor can be applied to compensate for the stretch. It is, of course, essential to calibrate the thermistor accurately through the range of temperature to be measured; one should also be aware that an extrapola- tion beyond the points of calibration may introduce As with the maximum thermometer, it is advisable to hold the resistance element a few feet above or below the point to be measured before recording the temperature. The current used in some sensitive single thermistors if applied steadily for a half minute may raise the tempera- ture of an element in air by as much as 0.2 °F and there- fore should be used only momentarily. The heating effect, however, depends on the thermistor and the current necessary to get the required sensitivity with the monitoring eqiupment. N o heating is observable with some of the more sophisticated equipment now available. In our work with a resistance thermometer, it was customary to measure the temperature at set points during the probe’s descent, leave the probe at the bot- tom for an appreciable period of time, and then repeat the measurements on the way up; at each check point we waited until equilibrium had been established. The two temperatures rarely diflered by as much as 01°F; the first temperature was influenced by the introduction of an element cooler than the surroundings and the second by the heat given off by an element warmer than the surroundings. The mean of two such measure— ments is assumed to be close to the actual temperature of the walls of the drill hole. If the resistance is measured at 5- or 10-minute intervals when the probe is in air and the readings are plotted against time, the equilibrium value is usually apparent well before it is reached. Temperatures were also measured with thermo- couples but this technique required that a cold junction be attached to the line no more than 20 feet from the thermocouple and lowered into the hole with it. If the cold junction was kept at the surface, the response be- came so sluggish when the hot junction was lowered a few hundred feet as to be unreadable even with the excellent portable potentiometer used. The volume and mass of the container used for the cold junction also discouraged the use of this device and led to the adoption of thermistors where comparable sensitivity was sought. Although the preceding discussion suggests many of the difficulties that arise in measuring temperatures in drill holes, one should realize that the temperature F4 measured may not be the temperature sought. If the drill hole contains a steel casing, the high conductiv- ity of the steel will do much to average the temperatures of the poorly conducting rocks through which the hole passes; furthermore the steel will carry the mean annual temperature wave much deeper into the hole than will the poorly conducting rocks. In cased drill holes in the East Tintic district, temperature perturbations caused by the annual wave were most perceptible in late summer and early spring. Convection currents rarely caused appreciable tem- perature changes in deep drill holes in the area under study, but sharp changes in barometric pressure did cause temperature changes corresponding to the com- pression or expansion of the air column in the hole. Where the water table is more than a thousand feet below the surface, compression of the column of air between it and the surface by a change in barometric pressure can cause a movement of as much as 25 feet in the column of air. A cloud shadow and the attendant cooling breeze passing over the collar of the drill hole one hot summer day was observed to cause a drop in temperature corresponding to such a fluctuation of the air column during temperature measurements made with a thermocouple at hole TS 14. It is generally assumed (Hales, 1937) that convection occurs in a water-filled tube if the water below a given point is less dense than the water at or above the point, except as inhibited by friction. Above 40°F, water expands with increasing temperature. The decrease in density downward due to the temperature gradient is opposed by the increasing compression caused by the weight of the overlying column of water. When the two opposing factors are equal, a critical gradient is said to exist. This critical gradient has been calculated by Hales (1937), but he pointed out that the basic as- sumptions must be in error because the calculated critical gradient is many orders of magnitude below gradients measured in geysers that erupt periodically. Hales suggested that the use of the molecular co- efficients for viscosity and thermal diffusivity in the final equation should be replaced by corresponding “eddy coefficients” which have been found to range from 1 to 400 as compared to a general order of magni- tude of 10‘3 for the molecular coefficients. Temperatures measured in water-filled holes having gradients in excess of 8"F per 100 feet clearly show that either convection does not take place or, if it does occur, it must be in many superposed cells of such short length that the temperature perturbations are to be measured in hundredths rather than in tenths of a degree or more. The minimum height of a cell is at least equal to the width of the tube and may be several times‘as great. SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY (Hales (1937) showed that the tendency toward con- vection varies with the fourth power of the diameter of the tube.) The drill holes in the East Tintic district for which temperature data have been obtained range in diameter from 2% to 6 inches and the tendency toward convection therefore is 33 times greater in the 6-inch hole than in the 2%-inch hole. For these holes the temperature differences for the minimum cell would be ——12(l0din.’ Where I‘ is the gradient in degrees Fahrenheit per 100 feet and d is the diameter of the drill hole; the differences are from 0.002 to 0.005 times the geothermal gradient. For a 6—inch hole with F: 10° per 100 feet, the minimum convection cell should then have a perturba- tion of 10X0.005 or 005° F. Perturbations that approximate this figure have been observed in some drill holes when the probe was first positioned but they commonly attenuated and disappeared after about an hour. An individual cycle of 005° F usually takes less than 10 minutes and the average of two cycles is Within :|:0.01° of the temperature ascertained by measure- ments prolonged over many hours. Rock temperatures that are accurately measured in a hole still may not give the true gradient. The tem— perature changes caused by drilling the hole diminish . slowly, and the longer the time spent in drilling the hole, the longer the time before the normal preexisting tem- peratures are regained. The true gradient is closely approximated long before the temperatures have re- turned to the predrilling condition. The minimum ‘ length of time that should elapse after drilling is com- pleted and before the temperature is measured should approximate the duration of drilling, but a period three times that long is desirable. Lachenbruch and Brewer (1959), reporting on an observation drill hole at Point Barrow, Alaska, stated that 67 days after drilling was completed (duration of drilling, 58 days) the gradient measured was only 5 percent greater than the probable equilibrium value, although the individual temperatures in this frozen terrain were about 2°C greater than the probable equi- librium values. These authors have devised a simple formula for computing the equilibrium temperature at any given depth when measurements are made at time intervals (Lachenbruch and Brewer, 1959, p. 105): temperatures are plotted as ordinates and the values of log, (1%) are plotted as abscissae, where t is time since drilling stopped—in either days or seconds—and s is the duration of drilling in the same units. Extrapolation ‘ of the nearly straight-line curve to a zero value for this expression gives the equilibrium value of the tempera- tures at the depth in question. The temperatures pre- dicted by this method for the Point Barrow hole after UNDERGROUND TEMPERATURES AND HEAT FLOW, EAST TINTIC DISTRICT, UTAH t=3s were within 0.05°C of those predicted by the use of 6 years’ data, when t=35s. The temperature at the bottom of an advancing drill hole is less disturbed by the drilling fluid than is that of the wall of the hole above. temperature gradient can usually be obtained if the bottom-hole temperature is measured about once a week during the time a hole is drilled, but the tempera- ture should not be taken until after a 24-hour shutdown. TEMPERATURES IN MINE WORKINGS The best time to get a temperature that approximates the temperature of the undisturbed rock in a mine is when the opening is first made; a side hole 4 to 5 feet deep drilled at the same time that the blastholes are made to advance the heading can be used to the best advantage if the temperature at the bottom of the side hole is measured within 24 hours. The high tempera- ture of the blast is only momentary and has no effect on the temperature of the rock 2 feet from the walls of the newly created opening, but the air used to ventilate the heading may have a perceptible efiect within a few days. As such prompt rock-temperature measure- ments are rarely available, the perturbations caused by mine openings must be recognized. Ventilation usually cools the rocks because of cooler temperature of in- troduced air and evaporational effects where rocks are damp or wet. Where freshly opened porous masses of sulfide are exposed to moving air, heating from oxida- tion of the sulfides usually proceeds more rapidly than cooling from ventilation and may cause a notable in- crease in the temperature of the rocks. The longer the period during which an opening is ventilated, the farther from the opening is the perturbation of temperatures perceptible. The difficult problem of calculating the original rock temperature has been treated mathematically by J aeger and Le Marne (1963) for the case where the temperature of the walls of the mine opening is held constant for substantial time intervals and the ventilation history of the opening is known. Such information was lacking for the mines in the East Tintic district. The calcula— tions and measurements made by J aeger and Le Marne show that in rock having moderately high diffusivity (a=0. 014), ventilation for about 5 years ha affected the rock temperatures to a distance of abou‘ 120 feet from the wall of the drift. l Where heating due to oxidation may be neglected, ventilation and moisture are the chief fadtors that change the temperature in the mine opening from that of the undisturbed rock. It is axiomatic therefore that temperature measurements should never be taken in a drift where air is moving unless this fact is recorded; 776—666—65—2 For this reason an excellent ' E5 temperatures so taken rarely correspond to those of the rock before the opening was made. Where mine open- ings disturb the pattern of subsurface drainage, a con- comitant change in rock temperatures is to be expected. Data from places in the East Tintic mines that were suspect were omitted. In mine openings that have been unventilated longer than they were ventilated, the rock temperatures a few feet from the opening may approximate those of the undisturbed rock; the conditions for equilibrium are similar to those discussed for drill holes (p. F2, F 3). Most drifts are floored with rubble commonly 6 inches or more deep on which the tracks were originally laid and such rubble, if dry, makes an excellent insulating medium. The conductivity of such material is usually only a fraction of that of the solid rock nearby and 6 inches of rubble may thus correspond in its insulating qualities to 2 to 3 feet of solid rock of the same composi— tion. A thermometer placed on the bedrock under the dry rubble in the bottom of a drift usually gives a temperature corresponding to that at the end of a hole 2 to 3 feet deep in the side of the drift. When mapping underground in the East Tintic district, U.S.G.S. geologists customarily excavated a hole to the bottom of the rubble in a drift, placed a thermometer at the contact of the rubble and the bed- rock, covered the thermometer with the material excavated from the bottom of the rubble, and then left the thermometer until it had reached a constant tem- perature. This temperature was then recorded on the mine plan with the notation as to whether or not there was an air current—“live air” or “dead air”. At any place in a drift where there was moisture, the tempera- tures were abnormally low because of evaporation even where there was no ventilation. Only temperatures taken in drifts that had long been unventilated are shown on the maps (pls. 3, 5). At the time of our study, all the accessible mine workings were above water table and were mostly unaffected by water circulation. Some temperatures from below water level in the Eureka Standard mine were measured earlier by engineers of the Tintic Stand- ard Mining Co. and may represent temperatures appreciably disturbed by ground water moving into the mine openings. In the mine workings where oxidizing sulfides are generating heat, the additional heat source is usually evident from the sulfurous odor in the mine workings as well as from the anomalous temperatures and gradients that are found near such areas. Temperatures recorded there rarely represent those characteristic of the undis- turbed rock and ore, and must be used with great caution. F6 TEMPERATURE GRADIENTS, CONDUCTIVITY, AND HEAT FLOW RELATION OF THERMAL CONDUCTIVITY, TEMPERATURE GRADIENT, AND HEAT FLOW The quantity of heat flowing through the rock is determined by both conductivity and gradient, and identical gradients can result in very different quanti- ties of heat moving through the rocks. In some areas in the East Tintic district the temperatures at water level are the same but, because of different conductivi- ties, the rocks above water level have markedly different values for the heat flow (q), as is shown on the map (pl. 1), southwest of the Eureka Standard shaft. The relation of the quantity of heat (q) passing through a uniform body of length (L) and area (a) for the time (t) is shown in equation 1, q=k(T1'ZT2)ta/’ (1) where k is the thermal conductivity and T1 and T2 are the temperatures at two points in the body separated by the distance L. The average temperature gradient I‘, through the distance L, is Tl—Tz F—T' <2) The amount of heat per unit area per unit time is therefore : q=ch‘. (3) In the United States, temperature gradients are generally specified in degrees Fahrenheit per 100 feet, but the conductivity of rocks is customarily given in cgs units (calories per centimeter per second per degree centigrade). Converting the gradient measured in degrees Fahrenheit per 100 feet to degrees Centigrade per centimeter requires use of the conversion factor shown in equation 4, q=lc1.823><10“I‘,, (4) where q is in cgs units and 1‘, is the gradient in degrees Fahrenheit per 100 feet. From equation 4, it is apparent that k in the cgs system (cal cm‘1 °C‘1) is Ic=q><104/1.8231‘,, (5) where q represents calories per second per square centimeter. If materials of different conductivities lie in horizontal layers and the lower and upper faces of the stack of layers are held at constant and different temperatures, we have conditions analogous to a simplified geological SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY structure. The temperature at the surface of the earth is assumed constant (equivalent to the mean annual temperature) and the temperature deep Within the earth is also assumed constant. If no heat source (such as oxidizing sulfides) and no heat sink (such as moving ground water) are present, the heat flow must also be constant, and the gradients in adjacent layers will be inversely proportional to their conductivities: g=k1F1=k2F2=knFn (6) where k1=k of layer 1, and I‘1=gradient in layer 1, and similarly for layers 2, 3 . . . n; hence P1_k_2. Where the conductivity of one rock layer is known, the conductivity of another layer of rock can be de- termined if the gradients in the two rocks have been measured. Where the layers are inclined rather than horizontal, the calculation of the conductivity may be less simple; but unless there is a steep dip together with a pronounced difference in the conductivities of the inclined beds, the conductivities determined by meas- uring gradients and using equation 7 give satisfactory values for k. (See pl. 2.) Roy (1963, p. 18; see footnote 1, p. F2) has derived an equation that permits calculation of the vertical gradient through inclined layers if the requisite param- eters are known. Roy assumed that a dipping tabular body of different [c from that of the surrounding media refracts the lines of heat flow: tan all k1 ——’ tan (12—1172 (8.1) where (11 and (12 are the angles with the normal of the incident and refracted heat-flux vectors, and [Cl and [Q are the conductivities. If end efl’ects are neglected and the isotherms in the underlying medium are assumed to be horizontal, then II sin2 al=(1—q—/1 —% ; 91 1 (8.2) Where a, is the angle of dip, g” is the vertical component of heat flow in the inclined tabular body, and q1 is the heat flow in the underlying medium. The ratio of flux densities is Q1 —=cos (12/ cos a1. 92 (8.3) For most places in the East Tintic district the calcu- lation of corrections for dip would be a meaningless refinement of the gross picture presented by calculations UNDERGROUND TEMPERATURES AND HEAT FLOW, EAST TINTIC DISTRICT, UTAH based on the simplified structure assumed, and there- fore no such correction was attempted. In our study of the East Tintic district, the conduc- tivities of many rocks were determined experimentally in the laboratory. The conductivities of many other rocks and formations were calculated by measuring gradients (see p. F10) and assuming that the conduc- tivities determined in the laboratory were appropriate for a particular rock type or geologic unit present in part of the section under study and represented by a sample on which laboratory work had been done. In many places the gradient at the boundaries of different units could not be determined, but only the average gradient for the inhomogeneous stack of layers. If the conductivities of the layers are known, however, it is possible to derive a figure for the average conduc- tivity of these rocks that is appropriate to use with the average gradient. Equation 6 may be written: JxTiJaTL . . . k_T', _ L1 ‘ L2 _ Ln (9) where T1,, Té, . . . T; are the differences in tempera- ture at the boundaries of the layers having conduc- tivities k1, k2, . . . kn. It is evident that [61:41:], (10) 1 and I__ qu, T,_——k1 (11) similar results being obtained for the other layers. The average conductivity (T) of a multiple-layered rock is (by analogy with eq 10) Tc=gd/(T;+T;+ . . . T1.) where d is the total thickness of all the layers, or (from eq 11) — L L qd ’ L392 9% k1 + 2 + + kn 01‘ - d [9:14 L2 7" (12) E+E+ r because the dimensions of L and d are the same, no conversion factor is needed if L and D are measured in feet and k in cgs units. The problem of multiple layers is common where a gradient is known between two widely separated points, F7 as at the surface and in a mine opening below, and it is desirable to know the temperature some distance below the point where the deep temperature was measured. If the strata in the section at the locality are known to be horizontal or to dip gently and if reasonable conductivities and thicknesses can be assigned to rock layers above and below the deep point (x), the tem— perature at a greater depth (y) is readily calculated from the heat flow and average gradient I‘z. By use of equation 8 the deeper gradient 1‘, is calculated from equation 14 (below) and the deeper temperature difference (TD is found and added to the measured temperature T,. q=kzrz=kyrw (13) where 7c, is average conductivity above point x and 7c, is average conductivity between points a; and y. Tc 1‘ F = f. I! 14 11 k, ( ) Tri=ruLm (15) and n=n+n no where TI is the actual temperature measured at the deep point, T, is the calculated temperature sought, and L, is the separation of y from 3:. Where temperatures could be checked at depth by‘ measurement, the temperatures predicted by the use of equations 12, 14, 15, and 16 were found to be in very satisfactory agreement with actual temperatures if the conductivities given in table 1 were used and if the geologic conditions approximated the simple configura— tions implied in the equations. In areas of complex geologic structure the results may be less satisfactory, and average gradients determined in steeply dipping rocks of very different conductivities give misleading results if the dip is unknown and assumed to be nearly horizontal. HEAT FLOW AND THERMAL GRADIENTS IN UNIFORM ROCKS, CONSTANT NONUNIFORM TEMPERATURE AT DEPTH Using the term “constant” only for time and the term “uniform” only for space, we note that wherever there is a constant nonuniform temperature distribution along a deep horizontal plane and the temperatures at the surface of the earth are essentially constant, the isotherms between the surface and the plane are sub— parallel to those of the surface at shallow depth and to those of the deep plane at deep levels. Where the horizontal gradient on this plane is nearly equal to the vertical gradient, the heat flux has a direction at F8 an angle to the vertical, and the amount of heat moving vertically differs with depth below the surface. The quantity of heat flowing vertically through the rock is almost the maximum flux if the horizontal gradient at the deep level is small. The maximum natural horizontal gradient observed in the East Tintic district is about 7°F per 100 feet in the Burgin mine and approximates the vertical gradient. By use of an electronic analog model simulator, as described by Karplus (1958), the approxi- mate dimensions of the region affected by a given distribution of isotherms at depth can be found. Assuming uniform conductivity, we first solved a comparatively simple problem: that of a change from a constant temperature of 140°F uniformly distributed north of a given line in a horizontal plane to 120°F uniformly distributed south of a transition zone 250 feet wide in the same plane, which is assumed to be 1,050 feet below the cooling surface (at 55°F). (See fig. 1.) The zone of appreciably disturbed isotherms above the temperature-transition area of the deep plane extends about 300 feet vertically and about twice this distance horizontally. The direction of heat flow in the homogeneous assumed medium would be per- pendicular to the isotherms, except that above the temperature-transition area of the deep plane and within 200 feet of it horizontally, the direction of maximum heat flow would depart from the vertical by less than 15°; the error introduced by assuming a uniform vertical heat flow thus would be less than 3.5 percent for a depth of 850 feet, and would diminish to less than 1 percent above a depth of 600 feet where the isotherms have a maximum deviation from the the horizontal of about 7°. Assuming vertical heat flow in areas containing appreciable horizonal gradients has inherent errors. The general efl’ect of ignoring these errors is to ShOW transition zones that may be wider than those that actually exist at water level and to average or minimize sharp thermal valleys and ridges. Although the general pattern of temperatures at water level is revealed by calculation based on gradients measured in the upper part of the section, details of the pattern in complex thermal areas can be determined only by measure- ments close to water level. In the second problem (fig. 2) solved by the electronic analog model simulator, the temperature profile approx— imates one that crosses the area of maximum tempera- ture contrast known in the East Tintic district. A tongue of relatively cool water lies between two broad areas of hot water in the western part of the Burgin mine; even there the distortion of isotherms a few hundred feet above the ground-water level is small, and SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY the narrow thermal trough representing the tongue of cool water would not be apparent from gradients measured more than 300 feet above the water table. The heat flow, however, appears to be approximately uniform along any vertical line except close to areas of sharp temperature changes at water level. Where evidence to the contrary is lacking, it has been assumed that the quantity of heat moving through the rocks above the deep source is constant for a vertical column a few feet in diameter, and that within a single vertical drill hole the quantity of heat moving up through the walls of the rock is constant unless the heat flow is modified by a current of air or water. Such conditions are usually indicated by the tempera- ture measurements. In the absence of such indications, differences in thermal gradients in a given drill hole can ordinarily be used with confidence to give con- ductivity ratios by use of equation 7 if due consideration is given to the depth of the hole, ground-water level, and possible horizontal temperature variations at water level. CONDUCTIVITY OF ROCKS DIRECT DETERMINATION OF CONDUCTIVITY The thermal conductivity of many rocks was determined by the method devised by Birch and Clark (1940). The amounts of heat required to maintain certain gradients in an insulated disk are accurately determined, and from these data the conductivity is calculated. Such measurements were made on many specimens from the East Tintic district (table 1). The laboratory measurements of “air-dry” samples made in Washington, D.C., in the early spring (1962) probably represent humidity of about 75 percent saturation and are closer to the measured conductivities of saturated specimens than to those of desiccated ones. The “air-dry” conductivities agree closely with the “in-place” conductivities for gradients measured above the water table; for gradients measured below the water table, however, laboratory values for saturated specimens are clearly appropriate. Apparently the laboratory values for “air-dry” samples closely approximate those of the rocks in their natural damp, but unsaturated, environment above the water level. INDIRE CT DETERMINATION OF C 0NDU CTIVIT Y The conductivity of rocks can also be determined from measurement of other thermal properties such as diffusivity (a): k a:—} GP (17) where a is specific heat and p is specific gravity. Meth- ods of determining the diffusivity of rocks in place UNDERGROUND TEMPERATURES AND HEAT FLOW, EAST TINTIC DISTRICT, UTAH .8 .13 .fl ocodm :2 25.3 .50: $8350 domauom wflwnfiaémflpnfioa E 323 038E059» :32? 32:28 2383809 .monon ~08 was .8: 30% Much maownowofion E Amos: Prawns mfiuoaeomm «o 53:95va mEBoAm «Edam S»£=§E%ofi.wo_aao_ 3536:3050 moomH \ z / “to: 3 03¢ co_um>w_w\._w>v_ 3521.9595 ”3.021 .003 kaONN \ “Town 005? KeONN .Oowwq are: come “:2: .88 loom .OOHm .OONm to O” .00 mm 5E 50% bomm “Tom u— mm .Oomm Sufism o 55ng 1313:! NI ‘SSENMOIHJ. H F10 from temperature measurements in shallow drill holes have been described by Lovering and Goode (1963). Given an accurate value for the diffusivity, the con- ductivity may then be determined by laboratory measurements of thermal capacity and density. Con- ductivities of some rocks in the East Tintic district were determined by measuring the specific heat and density in the laboratory and determining the difl’usiv— ity of the rocks in place by the shallow-hole temperature method described by Lovering and Goode (1963); this method requires two or more temperature profiles to a depth of at least 35 feet and taken more than a month apart. Where conductivities are known for certain rock units present in an uncased drill hole, the variations in temperature gradient within the hole SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY make possible the calculation, by use of equation 7, of conductivities of other rock units present (see pl. 2.) CONDUCTIVITY OF ROCKS AND FORMATIONS IN THE EAST TINTIC DISTRICT The “in~place” conductivity of the rocks in the East Tintic district ranges from about 1.7)(10‘3 for dry tuff to 13><10‘3 for wet quartzite. The presence of water in porous materials greatly changes the conductivity of the rocks, but in most of the areas investigated we were concerned chiefly with the conductivity of un- saturated rocks above the water table. The data for conductivity obtained by various methods for the rocks and most of the formations shown on the geologic map of the East Tintic district (Lovering and others, 1960) are shown in table 1. TABLE 1.—Thermal conductivity (k) of rocks and formations of the East Tintic district as determined by various methods Stratigraphic-unit symbols, after Lovering and others (1960): Quaternary units: Qal, alluvium: Qg, gr avel. . . . Tertiary units: Tpr, Packard Quartz Latite, Packard Rhyolite of Lindgren and Loughlin (1919); Tprt. basal tufi, rhyolite tufl of Lindgren and Loughlin (1919); Tao, Apex Conglomerate. Devonian unit: Dv, Victoria Formation. Devonian, Silurian. and Ordovician unit: DSOb, Bluebell Dolomite. Ordovician units: Of, Fish Haven Dolomite; Oop, Opohonga Limestone. Cambrian units: Ca, Ajax Dolomite; €op, Opex Formation; €cc, Cole Canyon Dolomite;€h, Herkimer Limestone; €d , Dagmar Dolomite; ere, Teutonic Limestone Co, Ophir Formation; €t, Tintic Quartzite. Method or calculation used to determine k: Sh hl, k calculated from diflusivities determined in shallow holes, according to method of Levering and Goode (1963). Lab , deter- mined experimentally. I‘p/I‘z, k calculated from ratio of gradients, k of one rock known; I}. gradient in rock named i111 column 3 (69 7); Pia/I'm in pyritized 1'00k; I‘d/1‘2. in dolomite; I‘Tprli‘z, in Packard Quartz Latite. Est, k estimated from comparison of lithology with that of rocks of known conductivity. Cale, k of formation calculated from conductivities and proportions of rocks comprising it. kx103: Number in parentheses shows number of separate specimens used in laboratory determination of k, or number of separate shallow holes in given material to obtain value of average k. None of the averages is more than $00003 from the maximum or minimum value of the group. [All laboratory determinations of conductivity are by E. C. Walker and William Huff, U.S. Geol. Survey, on waters of rock about 5mm thick cut from drill cores supplied by the Bear Creek Mining 00. Variation of k of different wet samples is such that the third digit is meaningless; therefore it is omitted] Stratigraphic unit Rock, alteration, weathering, and fracturing Thickness (feet) Symbol Method or calculation used to determine k kX103, in cal cm-1 °C-1 sec-1 Value of kX103 used Porosity (percent) Dry Air dry 1 Wet | Air dry 1 Wet 1 Qal 5—50 Alluvium and colluvium (sand and silt) ________________ Qg 5—50 Gravel (boulders, pebbles, sand, and silt) Tpr 2500+ Quafitz latite porphyry, unaltered ________ o ____________________________________________ Quagtz latite porphyry, plagioclase altered to calcite o ___________________________________________ Quartz latite porphyry, pytitized, unweathered_- 1 0 Quaatz latite porphyry, pyritized, slightly oxidized. 0 _____ o- _________________________________________________ Quartz latite porphyry, slightly argillized and iron- stained, weathered. Quartz latite prophyry, slightly arg‘illized and pyritized, weathered. Quartz latite prophyry, moderately sericitized, moder- ately weathered. Quartz latite porphyry, pyritized and argillized, moder- ately weathered. Quartz latite porphyry, moderately argillized, iron- stained, moderately weathered. Quartz latite porphyry, pyritized, argillized, strongly weathered. Do.. . _ Quartz lat a porphyry, strongly argillized, strongly weathered. Quartz latite tuft and breccia, calcitized, pyritized, and argillized. Quartz latite tufi, pyritized _____________________________ Pre-lava colluviuni, shale, and minor quartzite __________ Pre-lava colluvium, quartzite, and minor shale __________ Dolomite, limy dolomite, and minor sandstone, moder- ately fractured. 100 Dolomite and limy dolomite ____________________________ D 80 b 500 Dolomite , fresh _____________ Dolomite, leached (“sanded” Cherty dolomite __________ Shaly limestone ___________ Cherty and sandy dolomite ........ _ Dolomite, slightly leached and sandy ____________________ éee footnotes at end of table. 25—250 Tao 10—150 D v 280 Of 300 Oop 800 Ca 600 .v-su CO P9PPFP?PPN mmammmmmmw 1“ or 9’ S" P'PPS" w o ocoe UNDERGROUND TEMPERATURES AND HEAT FLOW, EAST TINTIC DISTRICT, UTAH F11 TABLE 1.—Thermal conductivity (k) of rocks and formations of the East Tintic district as determined by various methods—Continued Stratigraphic unit Method or kx10', in ca! m-1 "0-1 sec-1 Value of k)<103 used Rock, alteration, weathering, and fracturing calculation Porosity used to (percent) Symbol Th(lfckgess determine k Dry Air dry 1 Wet 1 Air dry 1 Wet l ee Cop 250 Shaly limestone and dolomite ___________________________________________________ 5. 6 ____________ (Dec 800 Dolomite, fine-grained _____________________________________________ 10. 0 ____________ Dolomite,nghtly leached and “sanded”- 6.9 ____________ 7.0 ____________ Ch 400 Limestone, 91 percent; shale, 9 percent .................. 5. 94 ............ 6.0 ____________ ed 80 Argillaceous dolomite ____________________________________________________________ 7. 5 ............ (its 400 Limestone .............................................. 6. 56 (l) 6. 76 (1) 6. 5 . 8 Hydrothermally dolomitized Cteequivalent to Sta llme- 10. 63 (2) 11. 85 (2) 10. 6 11. 85 stone tested. 60 Shale--.___ ______ -_- 3. 19 (2) 3. 58 (1) 3. 2 3. 6 col 80 Limestortle, 13 percent: sandstone, 12 percent; shale, 75 3. 61 ____________ 3.6 ____________ peroen . co 4 145 Limestone, argillaceous ............................ 5. 6 5. 6 ____________ Limestone, 63 percent: shale, 37 percent ...... 4. 38 4. 4 ____________ ct a 175 Dolomltg, 6 percent: sandstone, 17 percent; shale, 77 3. 62 3. 6 ____________ percen . Ct Quartzlte, slightly pyritlzed ............................. Lab ......... 1. 35 11. 83 12. 25 (2) 12. 25 13. 0 {2t 0 Quartzite, 90 percent: shale, 10 percent __________________ Calc ____________________ 9. 56 9. 6 ____________ ct 3000 Quartzlte, fractured .................................... Penn/Pg. .-_ ........................ 10-11 10—11 ____________ 1 Includes laboratory measurements made on air-dried samples, humidity about 75 3 U per part of Ophir Formation. percent, temperature about 80°F, and in-place conductivities of unsaturated rocks above 4 the water table. 2 Sample disintegrated during saturation. CALCULATION OF TEMPERATURES AT DEPTH The problems of establishing the temperature of a given point underground are of three kinds. If the temperature has actually been measured at the point in question, the uncertainty is reduced to assessing the actual technique of the temperature measurement and the extent to which the measured temperature repre- sents underground temperatures undisturbed by mining or drilling. A second type of problem occurs when the point for which the temperature is required lies well below points at which temperatures have been measured in different rocks and for which a temperature gradient has been calculated. Here the temperature gradient appropriate to the rocks where the temperatures were ‘ measured must be modified for any change in conduc- tivity of the rocks at greater depth. A third type of problem exists where the temperature is known at only one point underground and a calculated temperature is desired for a point at some specified elevation vertically above or below the place at which the temperature is known. This third problem can be solved with a fair degree of success if the conductivities 0f the rocks are known and if the mean annual temperature at the surface can be approximated. For many places in the East Tintic district where underground temperatures at only a single elevation above the water table have been obtained, the requisite quantities are known with suffi— cient accuracy to allow calculation of temperatures at the water table to within :l: 1°F. GEOLOGIG FACTORS Inasmuch as the calculation of temperatures below the point where a temperature has been measured de- pends not only on thegradients above this point but iddle pan of O hir Formation. 5 Lower part of 0p ir Formation. 0 Upper part of Tintic Quartzite. also on conductivity of the overlying and underlying materials, the geology must be interpreted correctly if the structure and type of material lying between the two points are to be correctly predicted. It is also im- portant to know the physical and chemical changes that have occurred in the rocks. The conductivity of a formation may change greatly with alteration—or may be very little affected. The replacement of feldspars in quartz latite by calcite (calcitic alteration) has no measurable effect on the conductivity of the “air-dry” rock; if, however, the feldspar is altered to clay, the conductivity of the rock may be decreased by 20 to 30 percent. Conversely, the change of a limestone to a dolomite may nearly double its conductivity. Al- though structural discontinuities such as faults and shear zones usually decrease the conductivity of a dry formation markedly, such zones when moist may have a much greater conductivity than the unbroken dry formation. In a section more than a thousand feet thick, however, such changes in conductivity have little effect on the average [C unless the fault zones are un— usually thick. In a region of complex geology such as is found in the East Tintic district, the calculation of tempera- tures at a given level must utilize the best geologic information available. Fortunately, the geologic maps and reports prepared by the mining companies working in the East Tintic district were made available for our study; these data, together with the experience we have gained in 20 years of detailed geologic mapping and petrographic study of the district, gave us, we believe, reasonably adequate understanding of the geologic factors in the places where the temperatures have been calculated at the water table. F12 SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY CALCULATION OF TEMPERATURES AT THE WATER TABLE As noted in the preceding section, the conductivities of rock types and formations have been measured, calculated, or estimated for the many rock units present in the East Tintic district, and correct cal- culation of temperature at water level as well as of heat loss depends on assigning correct conductivities to the formations present. The methods of calculating temperatures at various points are perhaps most clearly explained by the following examples that represent the three types of problems just noted. 1. Temperature measured at the water table in an abandoned drift: A temperature of 88°F was measured on the 1,450 (bottom) level of the Tintic Standard mine in the Tintic Standard-North Lily Development Unit at coordinates 31,890 N., 21,240 E. Ventilation had been stopped in this drift 2 years earlier, but the levels above it had not been worked for about 15 years. The ele- vation of the drift is 4,555 feet, about that of the water table; even though the temperature had been measuredwith a well-calibrated mercury maxi- mum thermometer, the figure obtained was never— theless suspect because it was appreciably lower than the temperatures measured in stub drifts nearby. The temperature Of 98°F measured in a raise 100 feet above the level indicated a reverse gradient. Temperature measurements were made nearby at higher elevations in the mine and the following temperatures were obtained: 67°F at an elevation of 5,604 feet, 98°F at 4,806 feet, and 98°F at 4,637 feet. The readings in mine workings above 4,555 feet probably approximated the rock temperatures except where they reflected the ventilation history of the drift at 4,555 feet. The reverse gradient between the top of the raise and the bottom level reflects the cooling caused by air circulating in the lower drift for several years prior to the terminating of exploration, which happened 2 years before our measurement was made. The problem then became one of pro- jecting the apparently “natural” gradient from the upper point (elev. 4,806 ft), where a tempera- ture of 98°F had been obtained, downward to the water level at 4,550 feet, the pertinent geologic conditions (see example 2) being considered. The temperature so calculated for water level (102°) was in harmony with those measured in the stub drifts a few hundred feet away. 2. Temperature calculated at depth from a gradient measured for a short distance below the surface: The Chief Consolidated Mining Co. diamond drill hole CC 71, about 2,800 feet northwest of the North Lily shaft, was drilled in porphyry of the Packard Quartz Latite many years before our tem- perature measurements were made, and presum- ably the temperature gradient represents the gradient in the surrounding rocks unafl’ected by the opening. The hole was caved at a depth of 370 feet, and only the gradient in the wet unaltered porphyry of the Packard Quartz Latite, above that depth, could be measured. However, the geology of this area is quite well known from mine work- ings in the vicinity and from other drill holes. The manner in which data were selected and cal- culations performed to solve this type of problem is illustrated in the following example. For calculation of temperature at water level (elev 4,550 ft) directly below drill hole CC 71 at 36,285 N., 18,540 E., the following information was available: Collar at 6,300 feet, bottom in porphyry of the Packard Quartz Latite at an elevation of 5,365 feet; water saturated from surface to depth of 367 feet (elev, 5,933 ft), Where the hole was closed by heavy ground. Data for calculation follow (values of k are times 103; abbreviations as in table 1): T2—T1_ 58.2—56.0 2.2 r (measured) : wl—xz _6050—5933—117 —1.897; thus, I‘zl.90°F per 100 ft and, taking [C from table 1, the values are— Tpr (wet), k=5.0, L1=6300~5933=367 feet, r= 1.9°, Tpr (dry), k=4.8, L2=5933—5270=663 ft (from surface and subsurface data), Tprt (dry) k=1.7, L3=5270—5220=50 ft (from thickness nearby), limestone (dry), k=6.5, L4=5220—5120=100 ft (geology from mine level at 5,105 ft), quartzite (dry), k=12.25, L5=5120—4550=570 ft (geology from mine level at 5,105 ft), T at 5,933 ft =58.2° (measured). From equation 12, the average conductivity projected from 5,933 feet to water table at 4,550 feet is I?><1o-3— 5933—4550 _663 50 100 6'03’ 570 = Zs+f7+fi+1225 and 12:6.03X 10*3 cal sec‘1 cm‘1 °C“. From equation 4, q=5.0><10‘3><1.9><1.82><10“4= 1.74 (from measured gradient in wet quartz latite porphyry), where q is calories per square centimeter per second. The average gradient between eleva- tions of 5,933 and 4,550 feet is given by UNDERGROUND TENIPERA'I‘URES AND HEAT FLOW, EAST TINTIC DISTRICT, UTAH f# q _ 1.743x1o-6 —1.82><10‘47c_6.03X10'3X1.823><10‘4 =1.583°F/100 ft. T at 4,550 ft=58.2°+ (1.58°><13.83)=80.1‘2°; .'.T,,,, the temperature at water level (4,550 ft), = 80°F. 3. Temperature at the water table calculated from only a single temperature measured above it in a mine opening: The water-level temperature was calculated for a position about 600 feet south- southeast of drill hole CC 71 discussed above. Temperatures measured at 50 feet, 150 feet, and 250 feet below the collar of CC 71 are respectively 53.7°F, 55.0°F, and 56.0°F, or I‘~1. 15° per 100 feet; by extrapolation of this gradient to the surface, the mean annual surface temperature at the drill hole is 56.0°F(——1.15><25.0)=53.1°F. ’l he isotherms shown on the map (pl. 1) ‘south of the collar of CC 71 suggest a slightly higher mean annual surface temperature, and an assurrled mean annual temperature of 54°F at the surface (elev. ., 6,315 ft), over the point of temperature measure- ment underground, should not be in error by more than a half degree. The underground temperature of 79.5°F was measured on the 900 level (elev., 5,110 ft) of the North Lily mine at coordinates 35,715 N., 18,850 E. (fig. 1). The rocks above this point on the 900 level are limestone, 70 feet (k=6.5); dolomite, 580 feet (k=7.5); and sericit- ized quartz latite, 555 feet (k=4.25). Below the 900 level the rock is quartzite to the water table 560 feet beneath. The average conductivity of the rocks above the level is given by 6315—- 5110 1205 3: k><10 555 580+_ 70 =.130 6+77. 3+10 4. 25 7. 5 =5.53, and ‘ k=5.53><10“3 cal sec"1 cm“1 °C“‘. The average gradient is (79.5—54)/12.05=2.116°F per 100 feet. q=—2. 12X5- 5X10‘3X1.8X10“=2. 13 me sec Mom where mc=microcalories (10" cal). The gradient in the quartzite is I‘— 2.13><10‘6 1. 8X10'4k_ 12. 25X10'3X1. 82X10'4 =0.95°F per 100 ft. (newsflash—3 F13 The temperature at water level T1,, is calculated to be: Tw=79.5+(0.95X5.60)=84.8, and T..=85°F. TEMPERATURE OF GROUND WATER GENERAL FEATURES As shown by the map (pl. 3), there is a great variation in the temperature of ground water which seems closely related to major structural elements such as the Eureka Lilly fault and the East Tintic thrust fault. The tem- perature of ground water at the water table in the main Tintic district has been measured. in only a few localities. Commonly it is consistent with a temperature gradient of 1.5 to 1.8°F per 100 feet, and with mean annual temperatures at the surface in the 40’s or lower 50’s. For example, the temperature in the Chief mine at water level (1,780 ft below the surface) at an elevation of 4,800 feet is only 71°F; a gradient of 1.5°F per 100 feet would require a surface temperature there of 44°F. On the eastern side of Tintic Valley, which lies just west of the East Tintic Mountains, the temperature in Mintintic diamond-drill hole 4, 2% miles south-south- west of Mammoth, is 741° at an elevation of 4,900 feet, 1,200 feet below the surface; the temperature gradient in this hole averages 1.67 ° per 100 feet and the surface temperature is 54°F. However, in Mintintic diamond— drill hole 2A, about half a mile farther south, the temperature was much higher, reaching 982° at an elevation of 5,190 feet, 700 feet below the surface. In hole 2A, however, the water was under strong artesian pressure and when tapped at a depth of 455 feet below the collar, it rose almost at once to within 20 feet of the surface. None of the mines in the main Tintic district are known to have had unusually warm workings even at- the greatest depths attained. We may say with confidence that the temperature of ground water is less than 80°F in the region west of the Eureka Lilly fault zone and west of its projected course south from the Eureka Standard fault system; the temperature at the Water table is in general that appropriate to its depth where the temperature gradient is 1.5° to 2°F per 100 feet. East of the Eureka Lilly fault zone, the tempera- tures at the water table are much higher than to the west and reach or exceed 140°F in several places. As shown on plate 4, the water table in the area of the monzonite stock south of Mammothstands close to the surface or within a few hundred feet of it; commonly it ranges in elevation from about 6,100 feet to 6,400 feet within the fractured stock but drops abruptly to about 4,900 feet in the Paleozoic rocks just north of the stock and then slopes northward to about 4,800 feet in the vicinity of Eureka. The ground-water surface also F14 slopes both east and west from this north-trending ground-water high, which underlies Eureka. To the east the water table falls at the rate of 45 feet per mile as far as the Eureka Lilly fault zone, beyond which its slope is much less, probably no more than 5 feet per mile for the next 2 or 3 miles. The configuration of the ground-water surface indicates that water is moving from the East Tintic Mountains westward to Tintic Valley and eastward toward Goshen Valley where the ground—water level—as indicated by Lake Utah, a large permanent body of water about 9 miles northeast of the East Tintic district—is at an elevation of 4,490 feet, only 60 feet below the water level in the Burgin shaft. The marked flattening of the water table east of the Eureka Lilly fault zone is in part caused by the greater fracturing and higher permeability of the rocks to the east but also may reflect the change in hydraulic grad— ient caused by the addition of water from underground springs east of the fault zone. EAST TINTIC THERMAL AREA The temperatures in the extreme southeastern part of the East Tintic district are higher than those found farther north (see pl. 3), but the thermal center to which they are related cannot as yet be defined. We will therefore attempt to describe the heat-flow quantita- tively only for the thermal area north of the Apex Standard thermal trough, which extends northeastward through the district a short distance southeast of the Apex Standard fault. This thermal trough crosses the Apex Standard workings about a thousand feet south- east of the Apex Standard No. 1 shaft. North of this thermal trough, and generally east of the Eureka Lilly fault zone, an area of at least 20 sq km (square kilo- meters) has a ground—water temperature at the water table above 80°F, but the position of the 80° isotherm in the eastern part of the district is not known. The average temperature at the water table for an area of 13 sq km, which includes the 80° isotherm to the west, is 104°F. For the area of 10 sq km that includes only temperatures above 90°F, the average temperature is 109°F. As shown on the map (pl. 3), the thermal max— ima show two distinct trends: a general north-south alinement along a north—trending thermal ridge, and northeast and southwest extensions from centers along it. Another striking feature of the thermal pattern is the subparallelism of cold-water areas to the northeast— or southwest—trending tongues of hot water. The linear thermal trends are strongly influenced by the geologic structures of the area. The northeast- trending tongues of hot water are evidently related to major northeast-trending mineralized fractures, and the transition from cool water west of the Eureka Lilly SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY fault into warm or hot water a short distance east is closely related to this major fracture zone. Not all the northeast-trending mineralized fractures, however, are marked by abnormally high temperatures. Both the Eureka Lilly fault zone and the North Lily shear zone are well mineralized in many places, but the ground water along these channels is not appreciably warmer in mineralized ground than elsewhere. It is also noteworthy that the north—trending thermal maximum (the Greyhound thermal ridge) which extends for about a mile northward from the Apex Standard No. 1 shaft through the Greyhound mining claims does not coincide with any structure observable in mine work- ings or drill holes in this area. The north trend, how- ever, is parallel to the strike of the vertical or over- turned beds and strike faults east of the East Tintic thrust fault and probably reflects permeable beds or fractures below the hanging wall of the East Tintic thrust fault, where the footwall of the thrust plate underlies the Greyhound thermal ridge. The highly permeable fault breccia of the East Tintic thrust fault also exerts a marked influence on the distribution of the hot and cold water, for in most places there is a sharp drop in temperature to the east of the thrust fault at water level. HEAT LOSS The heat loss at the surface above the abnormally hot ground water depends mainly on the conductivity of the rocks, the mean annual temperature, and dis- tance to the surface. It may, therefore, differ from place to place even where surface temperatures and ground-water temperatures are uniform. The general areas of greatest heat loss, nevertheless, do correspond with the areas of abnormally high temperatures at depth. Isograms through points of equal heat loss have been drawn for intervals corresponding to differ- ences of 1 mo cm‘2 sec"l (microcalorie per square centimeter per second) and are shown on the map (pl. 5). Inasmuch as the heat loss at the surface in the areas having ground—water temperatures less than 80°F corresponds in general to a measured loss of 1.5 to 1.9 mc cm~2 sec—1, we assume that any heat loss of more than 2 mc cm"2 sec”l is anomalous. On the basis of this assumption, the total anomalous heat loss in the East Tintic thermal area has been computed (table 2) and is found to approximate 265,000 calories per second for an area of 14.7 sq km. This heat loss amounts to 8.4 trillion (8.4><1012) calories per year. SOURCES OF HEAT Only two sources of heat for the abnormally high temperatures at water level seem possible: oxidizing UNDERGROUND TEMPERATURES AND HEAT FLOW, EAST TINTIC DISTRICT, UTAH TABLE 2.——-Heat flow in excess of 2 mo cm“2 sec‘1 in East Tintic area north of Apex Standard fault zone Mean excess heat Rate of excess heat Region (me) Area (cm?) flow flow (mc cm~z sec-l) (me sec-1) 2—3 44. 6X10g 0 5 22.3)(109 H 47. 1X10“ 1 5 70. 6X10“ 4-5 28. 0X10“ 2. 5 70. 0X10g 5-6 18. 3X109 3. 5 64. 0X10‘I 6—7 8. 35X10' 4. 5 37. 6X10° >7 .178><10g 5 0 .89X109 Total ....... 147x109=14. 7 km2 ____________________ 265x10°=265, 000 cal per sec sulfides and subterranean hot springs. During the early years of the present investigation, the heat gener- ated by sulfides in the Tintic Standard mine was very noticeable and a few years earlier had caused a mine fire which was brought under control only with diffi- culty. The temperature of the rocks was a matter of great interest for mining and ventilation engineers because of the practical problems of operating the mine; this interest and the possibility that abnormal geo- thermal gradients surrounding oxidizing sulfide bodies could be used in prospecting for blind ore bodies prompted the US. Geological Survey to begin its broader thermal studies. As the temperature data were accumulated over the years, however, they showed less and less direct relation to ore bodies, but until the Bear Creek Mining Co.’s extensive drilling campaign and their development of the Burgin mine, the tempera- ture data were insufficient to show a regional pattern. By the late 1950’s, however, the theory that areas of high heat flow always indicate the presence of sulfides at depth was becoming increasingly untenable. The hot saline water discovered on June 15, 1961, in the Burgin mine eventually provided the evidence necessary to establish the presence of subterranean hot springs; indeed, the first analysis of this saline water led Donald E. White, who had specialized in the study of thermal waters, to point out that this saline water was typical of a family of widely distributed surface saline hot springs in Utah, which had been sampled and analyzed in conjunction with his project. Several of the analyses shown in table 3 come from Mr. White’s project and are published here for the first time. The oxidation of sulfides in the main Tintic district causes no appreciable rise in temperature at ground- water level, and temperatures measured in the Chief mine in the vicinity of oxidizing ore bodies in limestone above the water table showed a maximum increase of only 3°F above the normal for the level at which temperatures were measured. Similarly in the North Lily mine, temperatures of 1° to 2° above normal were F15 found in partly worked stopes around small ore bodies adjacent to the Eureka Lilly fault zone and well north of the North Lily shear zone. Oxidation of sulfides in the San Manuel district in Arizona where the ore is undisturbed by mining caused an increase in the thermal gradient of 0.6°F or less per 100 feet (Lovering, 1948); if applied to the East Tintic district, such a change in gradient would increase the temperature at water level by less than 10°F, not enough to cause the temperatures of more than 100°F at ground-water level over large areas in the East Tintic district. Where sulfides have been disturbed by mining, the rate of oxidation may be sufficient to greatly increase the temperatures locally, but this increase could not possibly cause the pattern of thermal maxima in the vicinity of the Burgin and Apex Standard mines. No ore had been extracted from the Burgin mine at the time the temperature measurements were made and very little had been taken from Apex Standard mine when it was active 15 years before we measured the temperatures in it. Oxidizing sulfides, though capable of raising nearby temperatures a few degrees where undisturbed by mining, are com- pletely incapable of producing the high temperatures found in the East Tintic thermal area. If oxidizing sulfides cannot supply the heat, then it follows that most of the heat must be supplied by sub- terranean hot springs. This conclusion is also sup- ported by the composition of the water and the pattern of hot and cold tongues of water. Where the water table was first penetrated by an inclined winze in the Burgin mine, the water had a temperature of 139°F; a sample of this water contained 6,050 ppm (parts per million) of total solids. This sample was taken on December 26, 1960, shortly after the winze reached water level. Heavy pumping was necessary through the ensuing months while the winze was slowly ad- vanced. Water samples taken during the next several months showed a gradual increase in total salinity, which reached 7,480 ppm by November 1961 when the temperature was 138°F. During this period the chloride content increased from 3,090 to 3,670 ppm. The increase in total solids with heavy pumping strongly suggests a subterranean hot spring source, and the slight difference in temperature reflects only a slight perturbation probably caused by the cooling of the rock in the area of drawdown near the ventilated opening through which the water moved on its way to the sump. A sample from a drill hole 200 feet southwest of the winze sample showed a temperature of 128°F and con— tained only 72 ppm of chloride. Other samples in the tongue of cool water ranged from 780 to 1,560 ppm of chloride, whereas samples from the tongue of hot water F16 to the south proved it was also a high-chloride water which contained more than 3,500 ppm of chloride. The analyses show that the hot water in the Burgin mine is a saline water similar in composition to several surface hot springs in Utah (tables 3, 4). The striking similarity of the Burgin water to these hot-springs waters is especially evident when the ratios of some of the major elements are compared (table 3). The ratios SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY are nearly independent of dilution by fresh water or of concentration by evaporation, and thus emphasize the chemical relations that characterize this family of hot springs. There seems little doubt that the surface hot springs—represented by analyses 8 to 18—belong to a family characterized by high chloride, low to moderate sulfate, and low to moderate bicarbonate. The contrast between waters of this family and either TABLE 3.——Chemical analyses of water from Eureka city well, Burgin mine, [Analyses in parts Constituents a D t i ... a e o a Source collection ‘3 8102 A1 Fe Mn Ca Mg . Sr Ba Na K Li As Pb Zn 3 Ra U NH; H00; 804 1 Eurelllra city Dec. 31, 1940 44. 6 None _____ 59. 2 34. 2 ................. 7. 7 we . 2 North Lily Nov. 14, 1942 190 35 ___________ 340 112 mine. 3 Burgin 1111119.... Dec. 22, 1960 130 1.3 7 10.2 480 2240 1 0.36 ________ 11, 815 4 Burgin mine____ May 25, 1961 30 ________________ 315 62 ,700 5 Burgin mine__.. June 15, 1961 43 . 16 . 00 . 7 256 66 13. 1—0. 31 l0.3—0.03 2,000 6 Burgin mine... Aug. 28, 1961 _____ . 16 . 00 .00 396 39 ......................... 7 Burgin mine.... Oct. 1, 1961 48 .00 1 .64 375 53 15 1.4—.04 2,270 8 stinking June 8, 1954 48 1 05 .03 .00 946 297 31 4.1 12,600 Springs. 9 Cosoper Hot Prior to 1906 23. 5 0 0 ..... 267. 7 35. 1 ................. 591. 4 prings. 10 Rgosevelt Hot Sept. 11, 1957 313 <.04 < 04 .0 22 0 0 0 2, 500 prmg. 11 Crsystal Hot Oct. 27, 1951 31 ..... 61 .00 803 219 _________________ 14,700 763 8 ______________________________________ 465 466 prlngs. 12 Utsahflot Apr. 5, 1958 38 .46 .42 1.9 1,140 70 7,030 901 9 9 .00 01 00 39X10‘9 3X10“3 5.4 192 189 rings. 13 El onte Nov. 3, 1951 53 ..... .08 89 337 8 ................. 2,740 407 6 ______________________________________ 200 100 (Ogden) Hot Springs. 14 Hgoper Hot Nov. 3, 1951 28 .4 .01 1.7 523 118 2,390 283 2 ............... 100><10‘o .4X10-3 _____ 245 36 prmg. 15 Joseph Hot Oct. 11, 1957 84 . 1 . 56 41 264 44 __ l, 380 45 1. 5 .......... . 00 2. 7X10-9 1. 9X10‘3 ..... 412 l, 250 Spring. 16 Resd Hill Hot Sept 10,1957 54 .0 38 1 288 33 ................. 555 67 1 1 .19 ..... 10 __________________ 4 416 833 pring. 17 Agraliinam Hot Oct. 12, 1957 75 .0 0 .75 352 49 0 0 770 54 .0 .06 ............................ 1 142 704 pr gs 18 Bescks Hot May 19, 1942 32 8 ........... 653 134 4, 045 444 ......................... 119 875 PHI-183 19 Becks Hot Nov. 3, 1951 36 ..... .02 10 720 125 4,050 262 3 ...................................... 227 879 Springs. 1 Calculated from spectro aphic analyses by Kennecott Copper Corp. 2 Stearns, Stearns, and aring (1937, p. 183). 8 Concentration of radium~micromicrocuries per liter. 1. Eureka city drinking-water supply; composite sample collected 280 it from dug wells 2% miles northeast of Eureka. Collected by H. J. Hansen; analyzed by Utah State Board of Health. 2. Utah Count , face 01 1517 crosscut, 1500 level. Analyzed by International Smelting and Re ' g 00.; trace of free oxygen reported. 3. Utah County, 269 winze. Collected by J. D. Bush and Roger Banghart, Bear Creek Mining 00.; analyzed by Kennemtt Copper Corp. laboratories. 4. Utah Count , winze 269, 400 it down incline from 1050 level (70 it below water level). Collecte by T. S. Levering and H. T. Morris; analyzed by U.S. Geol. Survey Salt Lake City laboratory; lab. No. 23665. 5—7. Utah County, winze 269, 230 it down incline from 1050 level (20 it below water level). Collected by J. D. Bush and Roper Banghart, Bear Creek Mining Co. Chemical analyses y H. C. Whitehead (ab. Nos. 1517, 1536, 1525, respectively); spectrographic analyses by Harry Bastron and Robert Mays (lab. Nos. 62M11, 62M13, 62M12, respectively). 8. Located 6.8 miles northwest of Corinne, Box Elder County (White, Hem, and Waring, 1963, table 15, analysis 5). UNDERGROUND TEMPERATURES AND HEAT FLOW, EAST TINTIC DISTRICT, UTAH fresh shallow ground water or normal mine water is clearly shown by analyses 1 and 2 in table 3. Analysis 1 represents water from the perched water table in the gravels and the upper part of the lavas just northwest of Homansville Canyon in the area from which Eureka derives its water supply. The contrast in its ratios of NazCa and of 01:80.. and Cl:HCOa marks it as com- pletely unrelated to the Burgin water in origin. Al- F17 though the water from the North Lily mine (table 3, anal. 2) is dilute as compared to the chloride water of the Burgin mine, it contains far more sulfate and very little chloride and presumably represents water of the sort shown in analysis 1 after it had passed through oxidizing sulfides. The water of the North Lily mine is nevertheless relatively cool (about 80°F) in contrast to that of the Burgin mine. North Lily mine, and of saline hot springs of Utah having temperatures above 115°F per million] Constituents—Continued Temperature Dis— Ratios, by weight charge Tot 1 DH (gal De Ca Mg K Li 810’ goo. so. F Br B 1 a o o 1' _ _ __ _ - _. _. ._ _- _ Cl F B’ I N 01 N03 P04 B HIS Solids F 0 min) Na Ca Na Na $133; 01 01 01 01 01 1 53.0 0.1 _._. _._- _____ 4.0 ..... ---- -.-- 384 7.6 __________________________ 7 69 0.578 ............. 0.116 ....... 0.968 _______________________________ 70 -___ -.-- ---. _______________ ---. ---- 1,578 (acid) ............................... .318 ............. .12 _______ 13.4 _______________________________ 3,090 _-_. -___ _--- _______________ x6 _-_. 6,050 6.65 139 .50 0.07 0.006 (1)5 _______ .12 2,720 ---- -___ __-. _____ 2.1 _____ _-- --.. 5,500 6.6 127 .20 _____________ 005 0.24 .13 3,100 1.5.--- __-. 0.23 .0 0.31 6 0--.. 6,310 6.3 140 .26 .09 .0025 .007 .22 .12 3,590 _._- _._. --.- .00 .0 ..... --_- 0.0 ...... 6.4 138 .102 ................... .19 .11 3,670 1. 7 4.1 0.3 .00 3.1 .54 5.6 . 7,480 6.1 130 .141 100 0024 .006 .20 .11 . 21, 600 1.9 15 1. 3 . 00 . 0 3. 6 _-_. 60 6, 600 6. 7 2 122 . 31 045 00055 0013 . 013 . 0051 . 00009 0007 00017 6X10“ 634.6 _-_- _-_- _-_- ............... ---- _._- 1,794 ______ 144-156 62.2—68.8 2 100 .45 .13 . 06 .001 .01 ....... . 26 ............................... 4,240 7.5 3.3 .3 .01 11 2.2 38 _-_. 7,800 7.9 131 55 ’3 .009 <.04 .20 .011 .04 .04 .02 .002 0008 .900 7X10'5 25,000 .0 _-_- --_- ............... 4.5 .0. 42,200 6.5 132.4 55.8 1,320 .05 .37 . 052 .0005 . 0007 . 02 .02 <4X10-I ______ .0002 _________ 13, 300 3. 2 8. 2 2 . 00 0 . 00 5. l . 0 22, 800 7. 3 134 56. 7 I 20—25 162 . 061 138 . 0014 . (X127 . 01 . 01 . 0002 . 0006 0004 1. 5X10-‘5 5,060 3.4-.._ --.. ..... ..-.. ..... 3.6-..- 8,820 7.4 136.6 58.1 30:1: 12 .024 15 .002 .006 .04 .02 .0007 .0007 5,100 .6 .08 1.2-..- 8,600 6.8 139.8 60 30 .22 .23 .12 .0008 .003 .05 .007 .0001 ______ 0002 ......... 1,690 6.0 .04 4.8 _._. 4,970 6.6 147 64 25—30 .19 .17 .03 .001 .02 .24 .74 .003 ______ 003 --------- 660 3.0 0.3 0 ..... .0 .29 93.9 -.-. 2,705 6.4 168.8 76 50 .52 .12 .12 .002 .02 .63 1.27 .005 0005 .006 <10-4 1, 480 4. 5 1.8 l . 13 2. 5 . 06 . 9 ._-- 3, 560 6. 6 179. 6 82 25:1: . 46 . 14 . 07 <10—4 . 02 .10 . 47 . 003 . 001 0006 7X10" 7,668 ____ -._- _-_. _______________ _-_- -.-_ 13,872 6.5 -------------------------- .16 .21 .11 ________ .002 .02 .11 _______________________________ 7,260 2.3--_- ---- --------------- 2.6 7.513,500 ...... 132 55.5 ........ .18 .17 .06 .0007 .003 .03 .12 .0003 ...... 0004 _________ 9. Located M mile east of Monroe, Sevier County (Richardson 1907, $437). 14. Located 3 miles southwest 01 Hooper, Davis County. Analyzed by U. S. Geol. 10. Located 15 miles northeast of Milford on western slope of Mineral ountains, Survey Salt Lake City laboratory, lab.N 681. geaver. County. Collected by D. E.W11ite; analyzed by H. .Whitehead; lab. 0 11. Located 12 miles north of Brigham City, Box Elder County. Analyzed by U. S. Geol. Survey Salt Lake City laboratory; lab. No.7 12. Northernmost of 4 springs, 8 miles north of Ogden near boundary between gVeber and Box Elder Counties. (White, Hem, and Waring, 1963, table 16, analysis 13. Month 01' Ogden Canyon, Weber County. Analyzed by U. S. Geol. Survey Salt Lake City laboratory; lab. No.7 15. Located 1mile southeast of Joseph Sevier CountyI44 Collected by D. E. White; analyzed by H. C. Whitehead and J.P .:Schuch lab 0. 16. Red Hill Spring, Monroe Hot Springs,d1m11e northeast of Monroe Sevier County. Collected by D. E. White; analyzed bfig C. Whitehead; lab. 41.9 17. Near road north- northwest of NDelt2a, Jua County. Collected by D. E. Wlte, analyzed by J..P Schuch; lab.N 42.1 18-19. 4miles north of SalEt Lake Cit ,Salt Lake County. 18, collected by Leonard Tanner; analyzed by .McLach an. 19, analyzed by U. S. Geol. Survey Salt Lake City laboratory; lab. ENo. 7677 F18 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY TABLE 4.—Spectrographic analyses of residues of water samples from Burgin mine and Abraham Hot Springs [Analyses in weight percent; M, major constituent] Burgin mine, winze 269 Abraham Hot Springs Field sample Element (anal, table 3; date collected) 1 4121 I 4625 3 1517 3 1536 3 1525 3 D 3 (3; 2-22—60) (5—11—61) (5; 6—15-61) (6; 8—28-61) (7; 10—1—61) (17; 1963) Ag __________________________ 0. 0001 O. 0005 0 0 0 <0. 0001 Al ___________________________ . 005 0. 1 <. 0005 <. 0005 <. 0005 0 As __________________________ . 02 . 10 0 0 0 . 3 B ___________________________ . l . 1 005—. 05 . 05 . 005—. 05 . 001 Ba. __________________________ . 008 0 0005—. 005 . 0005—. 005 . 0005—. 005 >. 5 Be __________________________ . 00008 0 0 O 0 . 005 Ca __________________________ 5 22 M M M 3 Cb __________________________ . 001 0 __________________________________________ 0 Cd __________________________ . 005 0 0 0 O <. 002 C0 __________________________ . 000 . 0001 0 0 0 <. 0005 Or __________________________ . 0005 . 0007 0 0 0 <. 0005 Cu __________________________ . 002 . 0004 <. 0005 . 0005 <. 0005 . 0005 Fe __________________________ . 02 . 3 . 0005 0005 . 0005 >20 Ga __________________________ . 0004 0 0 0 0 . 003 Ge __________________________ . 0001 0 0 0 O <. 002 In ___________________________ . 00007 0 0 0 0 0 K ___________________________ 2 2 __________________________________________ 0 La __________________________ 0 0 __________________________________________ . 005 Li ___________________________ . 2 . 5 0 0 0 ______________ Mg __________________________ 1 4 . 05—. 5 . 05—. 5 . 05—5 . 15 Mn __________________________ . 003 0 <. 0005 . 005 <. 0005 >. 5 Mo __________________________ . 00005 0 0 0 O <. 0002 Na __________________________ 30 18 __________________________________________ 0 Ni __________________________ . 00008 . 0000 0 0 0 <. 0005 P ___________________________ . 05 0 0 0 0 0 Pb __________________________ . 003 . 007 0 0 0 >. 1 Pt ___________________________ 0 0 Tr <. 05 0 0 Rb __________________________ . 006 0 ________________________________________________________ S ____________________________ 4 4 ________________________________________________________ Sb __________________________ . 02 0 0 0 0 . 15 Sc ___________________________ 0 0 0 0 0 <. 0005 Si ___________________________ . 5 . 7 05—. 5 . 05—. 5 . 05—. 5 0 Sn __________________________ . 0003 0 0 0 0 <. 001 Sr ___________________________ . 06 . 05 05—5 05—. 5 . 05—. 5 . 5 Ti ___________________________ . 0007 0 0 0 0 . 02 V ___________________________ . 0006 . 0008 0 0 0 . 003 W ___________________________ 0 0 0 0 0 . 5 Y ___________________________ . 002 . 002 __________________________________________ . 0015 Zn __________________________ . 007 . 1 0 0 0 . 02 Zr ___________________________ . 001 O 0 0 0 <. 002 IAnalyst, R. E. Word, Kenneoott Copper Corp. Looked for, but not detected: Au, Bi, Ce, Dy, Er, Eu, Hf, Hg, Ho, Lu, Nd, Os, Pd, Re, Rh, Ru, Se, Sm, 'l‘a, Tb, Te, Th, T1, Tm, U, Yb. Table 5 shows the chloride content and temperatures of nine pairs of water samples taken from the same points in the Burgin mine about a month apart; although most pairs of samples are nearly the same in both temperature and composition, a few show striking differences which must be attributed to the movement of chloride-rich waters through dilute meteoric water. The decrease in temperature of the water samples taken from the inclined winze may be due to the cooling of the rock during mining operations, but the increase in chloride content can only be explained by the intro- duction of more concentrated brine. The h » avy pump- ing as the winze progressed probably cause deep hot- springs water to move through the fractured rocks that 2Analysts, Harry Bastron and Robert Mays, U.S. Geol. Survey. Looked for, bigt not detected: Au, Bi, Hf, Ir, Nb, Pd, Ta, Tl. Analyst, Maurice De Valliere, U.S. Geol. Survey. precipitate collected by W. R. Griflitts; lab. No. 63—108 8. SPRINGS 23ample of black ferruginous had previously been in equilibrium with the tongue of cold water just south of the winze. All the evidence accumulated by the end of 1963 supports the conclusion that the centers of highest temperature reflect the circulation of thermal water which rises along a north-trending fracture zone in the footwall of the East Tintic thrust plate and spreads northeastward or southwestward trending fractures in its hanging wall. (See fig. 4). along northeast- APPROXIMATE FLOW OF SUBTERRANEAN HOT Although the amount of hot water supplied by the hot springs cannot be directly measured, the flow can UNDERGROUND TEMPERATURES AND HEAT FLOW, EAST TINTIC DISTRICT, UTAH F19 TABLE 5.—Variation in chloride content, temperature, and relative deuterium content of water samples from Burgin mine [Water samples collected April 11 (1) and May 16 (2) 1962 from drill holes {mm 1,050 level] US GS field No. Chloride (ppm) Temperature (T) Deuterium analyses Drill-hole (May 16 samples) Location coordinates No. or location 1 2 1 1 2 2 1 2 Lab. No. D : H (percent) 3 401—TL—623 401—TL—62m 30,392 N., 26,170 E _________________ ET 71 564 510 120 120 ____________________________ 402 2 31,160 N., 26,040 E ..... __ B 29 1, 380 1, 390 139 138 ____________________________ 403 403 30,900 N., 26,720 E ..... . 269 winze 3,560 4 3 480 139 119 I.F. 3160-17 —13. 1 404 404 31,100 N., 27,433 E.. _ ET 26 3, 440 3 200 110 110 ____________________________ 405 405 30,425 N., 26,900 E_- . B 8 1, 560 2, 040 125 123 ____________________________ 406 406 30,717 N., 26,485 E.. ET 80 73 80 128 128 LF. 3160-16 —10. 9 407 407 30,535 N., 26,697 E.. B 33 2, 290 2, 430 126 1 ____________________________ 408 408 30,713 N., 27,715 E ________ B 32 3, 520 3, 590 135 137 I.F. 3160-19 —11. 0 409 409 30,825 N., 25,862 E _________________ B 24 780 4, 070 139 139 I.F. 3160—18 —-12.4 lAnalysis by K. E. Edwards, U.S. Geol. Survey. ’ Analysis by Kennecott Research Lab., Salt Lake City, Utah. 3 Isotopic analyses by Irving Friedman, U.S. Geo]. Survey; the figures show relative (lgifiegpvm content of the sample compared to that of the standard mean ocean water be estimated Within certain limits if we know the amount of heat escaping at the surface and the amount of hot-springs water required to raise the temperature of the meteoric water to that Which has been found. As noted earlier, the heat loss for much of the East Tintic thermal area can be computed from the data shown in figure 6, and in turn, the heat loss‘ that is above normal can be equated with the amount of hot water necessary to supply this heat if the difference in temperature between the subterranean thermal water and the cool meteoric water is known. Assuming that the average temperature of ”normal” ground water at the water table is about 80°F (26.67°C) and that the maximum temperature at the water table of the rising thermal water is 143°F (61.65°C), we have a tempera- ture difference of 63°F (35°C). The total anomalous heat flow at the surface in the area studied (pl. 5) is approximately 265x103 calories per second (table 2); it represents heat from a quantity of water 35°C hotter than the normal water-table temperature. This water has 35 calories per gram more than does water at 80°F (26.67°C), and the total excess heat flow is thus equivalent to an influx of water at 143°F of about 2——653>5<103grams per second, or 7.6)(103 grams per sec- ond—approximately 2 gallons per second. The amount of hot water required to heat the cool ground water is more difficult to estimate but is a comparable quantity. An approximation of the amount of ground water moving through the thermal area must first be made; the general slope of the water table and the control of ground-water movement by major frac- tures are known. (See pl. 4.) Because of the accel- erated evaporation and limited rainfall during most of the hot summer months (June—September), no appreci- able amount of precipitation reaches the water table; however, during the Winter and early spring When runoff, evaporation, and transpiration of water by vegetation (%) percent SMOW= MAW—mince (3) are minimal, an appreciable part of the total precipita- tion probably moves down to the water table and slowly migrates away from the hydrologic divide that passes under Mammoth and Eureka. The precipitation at Eureka, just west of the East Tintic area, and at Elberta, a few miles to the east, averages about 15 inches and 10% inches, respectively. (See table 6.) Quantitative data on rate of ground-water recharge from precipitation in areas near the East Tintic dis- trict are lacking. Recently, however, Gates (1963) estimated water loss in the Oquirrh Mountains, which are just north of the Tintic Mountains. Evapotrans— piration is the greatest single factor in water loss. Although Gates did not have exact figures for trans- piration losses in the Oquirrh Mountains, he utilized the quantitative figures obtained by Croft and Mon- ninger (1953) in the Wasatch Mountains east of Farm- ington, Utah, Where 44 percent of the total precipitation is transpired in areas covered by aspen and herbaceous vegetation. Unfortunately these figures apply chiefly to steep mountain watersheds between 7,000 and 10,000 feet in altitude and are not directly applicable to the Oquirrh Mountains. Gates, however, noted that the percentage of precipitation represented by evapotrans- piration increases as total precipitation decreases and as average temperatures rise. He therefore concluded that two-thirds, or 67 percent, of the precipitation is lost by evapotranspiration in the Middle Canyon area of the Oquirrh Mountains. Adding to this figure the measured water losses from surface runoff, channel underflow, springs, drains, and wells, Gates found that approximately 16 percent of the total precipitation is unaccounted for and can be presumed to be leakage to permanent ground water. The average precipitation in the Tintic area is about half that in the drainage basin of Middle Canyon, and the proportion of water lost by evapotranspiration thus should be substantially greater there than in the Oquirrh Mountains. Because 4Sample taken May 11, 1962. F20 SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY TABLE 6.—Prect'pitation and ground-water recharge in the East Tintic district [Data in inches; precipitation data item U.S. Weather Bur. (1952, p. 189)] Elberta Eureka Ground-water recharge, East Tintic thermal area Month Departure Departure 20-yr average 1950 precipi- from 20-yr 20-yr average 1950 preclpi— from 20-yr In Paleozoic In lava precipitation tation average precipitation tation average bedrock 1 bedrock 2 precipitation precipitation 0. 81 1. 61 +0. 80 1. 36 2. 32 +0. 96 0. 136 0. 068 1. 02 . 42 —. 55 1. 52 . 70 —. 82 . 152 . 076 1. 01 .38 —. 73 1. 47 . 66 —. 81 . 147 .073 1. 02 . 47 —. 55 1. 34 . 17 —1. 17 . 134 . 067 1 15 1.02 —. 13 1. 36 .75 —. 61 .068 .034 59 .09 —. 50 .92 .33 —. 59 ____________________ 83 . 88 +. 05 98 1 48 +. 50 ____________________ 79 .04 ——. 75 1. 38 12 —1. 26 ____________________ 71 .64 —. 07 .62 53 —. 09 ____________________ 99 .38 —. 61 1. 59 . 33 —1. 26 . 080 . 040 73 1. 02 +. 29 . 96 1. 30 +. 34 . 096 . 048 74 .48 —. 26 1. 58 1. 02 —. 56 . 158 . 079 10. 39 7. 48 —2. 01 15. 08 9. 71 —5. 37 . 971 . 485 l Estimated from 20-yr average at 10 percent of precipitation during spring, late fall, and winter and at 5 perwnt during May and October. the East Tintic area has higher temperatures and less total rainfall, it seems reasonable to assume that evapotranspiration there is closer to 75 percent than to 67 percent and that, accordingly, no more than 5 to 10 percent of the total precipitation reaches the water table. If we assume that 10 percent of the total rainfall during the late fall, Winter, and spring reaches the water table in the East Tintic Mountains, we may assign no more than about 1 inch (2.5 cm) of the precipitation near Eureka to ground-water recharge. Much more of the water falling on the Paleozoic bedrock surfaces is available for deep water recharge, however, than is available from the rainfall on the lava terrane. Perched water tables under gravel are common on lava but are rare in the Paleozoic rocks. The lava is rela— tively unfractured, but the Paleozoic bedrock is highly fractured, folded, and stratified and thus is well suited for conducting moisture to depths; it is assumed arbitrarily that in the lava terrane half as much moisture reaches the water table as in the areas of Paleozoic bedrock. The map (pl. 4) shows the areas of Paleozoic bedrock and lava and also the approximate elevation of the water table. It is further assumed that recharged ground water moving through the East Tintic thermal area is derived from precipitation in the area that lies east of the hydrologic divide, west of the eastern part of the Greyhound thermal ridge, and between the Apex Standard and the Centennial—Homansville Canyon fault zones, and that the ground water moves generally toward the northeast and east. The area assumed to supply ground water is shown by pattern on the map (pl. 4). 9 Estimated as half that in Paleozoic bedrock. The recharge area includes 11.2 sq km of Paleozoic bedrock and 19.4 sq km of lava terrane. Com- putation—((2.5><11.2><101°)+(1.25x19.4><101°))/103— shows that the annual increment of water available from precipitation in these areas is in the order of 52X 107 liters per year, equivalent to about 16 liters per second. Of the recharge area, 13.2 sq km is underlain by ground water that has a temperature in excess of 80°F; the average temperature at the water table for this ground water is computed as 104°F. To raise 16 liters per second of water having a temperature of 80°F to an average temperature of 104°F by addition of water having an initial temperature of 143°F requires about 6 liters per second or approximately 100 gallons per minute. If the initial temperature of hot-springs water was higher or the recharge was less than assumed, the volume of hot water required would diminish proportionately. The heat loss from the rock cover and the heat required to raise the temperature of the incoming ground water would be supplied by the sum of the two quantities of water at 143°F whose equivalence in heat energy has been calculated (above and p. F19); the sum is 3.7 gallons per second or 220 gallons per minute. After allowing for the uncertainties entering into the calculations, we can say that the subterranean hot springs in the area north of the Apex Standard thermal trough must have a flow of no more than a few hundred gallons per minute. ORIGIN OF THERMAL WATERS The isotopic analyses of the chloride brines (table 5) by Irving Friedman show deuterium-hydrogen ratios UNDERGROUND TEMPERATURES AND HEAT FLOW, EAST TINTIC DISTRICT, UTAH (D :H) that are completely different from those charac- teristic of ratios in undiluted connate waters. The D:H ratios for normal connate water are commonly between 0.0 and 4 percent less than for standard mean ocean water (SMOW). The ratios of the Burgin water samples range from —10.9 percent for the most dilute water (73 ppm 01), to — 13.1 percent for the most concentrated brackish water analyzed for deuterium (3,560 ppm 01) in the sample from the inclined winze; such ratios are appropriate to either magmatic water or meteoric water, but would also be found in any water that had been greatly diluted by either magmatic or meteoric water. Tenfold dilution of a connate water of a 0.0 D :H percent would change the ratio of meteoric water having —12.5 D:H percent only to —11.2 D:H percent. Although the determinations do not suggest that the negative values of D :H percent increase with dilution, the variations are erratic and could represent mixed waters or the effect of local evaporation. Tritium analyses of some typical waters of the area furnish additional data bearing on the origin and length of residence of the ground water. Background informa— tion: tritium has a half life of 12.5 years; the average for the Mississippi River before the first hydrogen bomb (1954) was 5 T.U. (1 T.U.=1 tritium atom per 1018 hydrogen atoms, ~32 picocuries per liter), in 1955 the average rose to 44 T.U., and in 1959 to 119 T.U. (Junge, 1963). Samples were collected in June 1964 and analyzed for AD/H (SMOW) by Irving Friedman and for tritium by G. L. Stewart 6 months later. A sample from the hottest of the Abraham Hot Springs, 50 miles west of Eureka, temperature 185.5°F, con- tained 14.8 T.U. (AD/H2138). In contrast, the cool Aperdue Spring 4,000 feet southwest of the Apex Standard N o. 1 shaft, temperature 60°F, contained 445 3:45 T.U. (but AD/H=12.6) and probably repre— sents shallow perched ground water fed by the current annual precipitation. Deeper perched water tables tapped by the Newmont No. 2 shaft between 190 and 700 feet below the surface, flowing at 20 gpm, tempera- ture 86°F, contained only 12:1:3 T.U. (AD/H='12.6). F21 A sample of acid mine water from a winze used in mining a few months earlier—the 261 winze of the Burgin 1,050 level—had a temperature of 104° F, and contained 25 :1:4 T.U. (D/H=12.5). The deep drill hole at coordinates 30,645 N, 26,598 E, about 20 feet west of the N0. 2 vertical winze on the same level of the mine represents ground water undisturbed by mining or drilling for at least 2 years; its temperature was 133 °F and it contained 7.5 T.U. (D/H=13.2). A general decrease in tritium content and therefore an increasing age is evident with increasing temperature and depth but, unfortunately, adequate samples of the hot water (140 °F) first tapped by the inclined winze of the Burgin Mine were not available. The chemical character of the waters of many thermal springs yields more persuasive evidence of their ultimate origin. Much attention has been given to this problem by White, Brannock, and Murata (1956), White (1957), and White, Hem and Waring (1963), but here it is only necessary to focus our attention on a few components of the saline water. The ratio of bromine to chlorine in brines furnishes one of the best indices of marine or connate water as contrasted with meteoric waters that have dissolved marine evaporites, but the ratio does not adequately distinguish between magmatic water and saline lake water of closed basins. Valyashko (1959) made an excellent study of the theoretical and practical aspects of the distribution of bromine in brines and of the salts that crystallize from them. He found that the distri- bution coefficient between solid and solution for bromine included in the crystal lattice of halite separating from a bromine-bearing brine had a ratio of nearly 30:1; the bromine content of the salt crystallizing from the brine was thus only a small fraction of the amount left in the brine. Some selected figures from tables prepared by Valyashko are given in our table 7, which shows that the ratio of bromine to chlorine in normal marine con- nate brines is greater than 0.0032, whereas in the halite the ratio is less than 0.0032 by at least one order of magnitude; the BrzCl ratio for the brine of Great Salt TABLE 7.——C’hlorine and bromine data for evaporating sea water and the resulting solid phases [Data from Valyashko (1959, tables 2, 5)] Brine (percent'chlorine Ratio (Bl-X103 ol bromine to chlorine in— and bromine) Cl Specific gravity Cl Br Brine Halite Sylvite Carnallite 1.01 _________________________________________________ 1. 01 0.0035 3. 48 ______________________________ 1.199 ________________________________________________ 14. 33 .047 3. 28 ______________________________ . 15. 72 . 051 3. 24 0. 11 ____________________ . 15. 77 . 226 14. 30 ______________________________ . 15. 93 . 236 14. 80 . 44 2. 9 __________ 1. 323 ________________________________________________ 19.10 . 334 17. 50 . 61 4. 00 __________ 1. 325 ________________________________________________ 19.23 . 342 17. 80 . 60 4. 20 8. 3 F 2 SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY j ”vii???“ HOT SPRINGS ASSOCIATSED CONNATE [N VOLCANIC GEVSERS AND BOILING HOT SPRINGS IN UTAH; BURGIN WiTH BRINES AREAS; SPRINGS IN TEMPERATURE MINE sum: TEMPERATURE VOLCAN'C AREAS >125-r <180°F WATERS DEPOSITS <1“ F :3 27 '29 -_ -- 25. . ' '28 I- _- 26 _ 9 21.2.3' '24 . 020 22 “f 17. 018 40 :: O ._ -15 16 o -_ D __ __ 14 039 __ x 370 °38 I m 0101—:— 13 44 1: E :: X12 036 g, o :: ’0‘ __ X o 942 __ m —— 11 35 1 __ " " ' 32 34 ° . N 4 9 10 XXX " a " X x 33 ~~ 19- x N 0001:.— 31 __ 9 " 8X 30xx 35 m :: -- -_ 2 __ “ X :: _. 7 -— 0.0001 —__— __ :5 X6 25 I: X1 5 I :: >< -- 000001 %__“____J__ W Wflfi— 0'01:— 5 .24 ‘:_— E X 11 E 33 2108140" 25 :; —— x 6 28 19 '— __ o _ o 13 22' ' 18 2‘5 0001?? X7 X 400 039 42 .: E _x"31 :: __ 38 32 I: _- 4 .17 _ ._ 1x>< , 21 - 0.0001—:x .25 ___ T—“_—%—— x 5 1'5 —__~ —~ ““ 13 :: -_ >f2 24. 32 o4 _ 14 X 20 . '25 fl 0 1__ 1; 15 21. ° 17 3335a .g§°36 44 “5 ‘ SE 3 . 25 1,8" '29 22 a 43 o 3— z :: x 10 23 .' 27 30 x 39 _, .- 1 x 34 o —» M " x1 9 x 23 1 42 v " V X 11 37 a 4 8 ~- “ x . 001E: 6 :: :: X7 I 2 __ ->< 0.001 FIGURE 3.——Ratios of Si02 to total solids, BrzCl, and KzNa in subsurface waters. Heavy line shows mean of group; dashed line shows mean of group including sample 5, which was collected from a potash salt mine. 0, fresh water (less than 3,000 ppm solids); O, brackish water (3,000—10,000 ppm solids); X, brine (more than 10,000 ppm solids). UNDERGROUND TEMPERATURES AND HEAT FLOW, EAST TINTIC DISTRICT, UTAH F23 Samples shown in figure 3 Temperature Sam- Locality Reference Analy- ple sis °C °F 1 Brine seep near Malaga, Eddy County, N. Mex ________________ 37 _____________________ 98.6 ___________________ White, Hem, and Waring (1963, table 9 2 Salt spring, 7 miles north of Newcastle, Wyo. Water from the 9 ______________________ 48.2 ___________________ Gott and Schnabel (1963, table 2) ________ 2211 Minnelusa Formation. Auger holes in Great Salt Lake Desert, Tooele County, Utah; Unknown _____________ Unknown _____________ White, Hem, and Waring (1963, table 4 composite of 126 brine samples. Brine (oi-med by meteoric 27). water that leached saline clay and sand seams a few feet deep. Glenwood Springs, Garfield County, Colo ___________________________________________ 125.6 _______________________ do ___________________________________ 6 Brine seep in National Potash Co. mine, Lea Count, N. Mex.. Unkno own _____________ Unkno __________________ do ___________________________________ 7 uerm'nes 87. 5 (bottom of hole); 189. 5 (bottom of hole); White, Hem, and Waring (1963, table 3 Parish,L 42 (discharge). 107 6 (discharge). 13.) Gulf Oil (13:10:!)- well PP-19,Timba1ier Bay oilfield, Latourche 70 Hem, and Waring (1963, table Parish Seaboard Oil Co. well S. ’1‘. U. 305-13, Fresno County, Calif. --_- Amerada Petroleum Corp. Peters gas well, Maine Prairie, Solano County, Calif. 10 Well, near Niland, Imperial County, Calif 40 4 2 Gulf Oil Corp. well 28—E, West Bay 011 field, P 7 8 9 Hem, and Waring (1963, table Hem, and Waring (1963, table Hem, and Waring (1963, 11 Spring, Kuau-Tsu— Ling, northern part of Taiwan 12 Springs, Cerro Prieto, Baja California, Mexico--. 13 Sprudel spring, 20 miles west of Mount Vesuvius, Italy .......... 14 Spring 1, Nalachevskie grou apb southeast Kamchatka, U. S. S. R.- 15 Spring near Jemez, Sandov County,N Mex .................. 6 16 Tazuniataipuhipuhl Geyser, Tokaanu, North Island, New ea an . 17 Growler Spring. Morgan, Tehama County, Calif 18 Paryashchii 1, Pauzhetsk, Kamchatka, U. S. S. R ________________ 19 “Geyser" spring, Mono County, Calif- ................. 20 Drill hole 4, Wairakei, North Island, New Zealand.-._ 21 Vpring 8, Steamboat Springs, Washoe County, Nev... likan Geyser, Shumhaya, Kamchatka, U. S. S. R 23 Geyser H-1,Umnak Island, Alaska _____ 3 _ 228 (at 970—ft depth). -. £29 (at 379-ft depth). _ , H ,4 White, Hem, and Waring (1963, table 18)- Whitle), Hem, and Waring (1963, table 17)- H 199. 442. 4 (at 970-ft depth;- 343.2 .6 (at 379- ft depth H 24 Sugar-bowl Geyser, Kamchatka, U. S. S. R-. p—i 25 Spring, Upper Basin, Yellowstone National Park, Wyo_.-_-_--: 94 26 Spring, Norris Basin, Yellowstone National Park, Wyo__- - 27 Spring, Beowawe Geyser-s, Eureka County, Nev Ssjodangi spring, south of Geysir, east-northeast of Reykjavik, Icel an ‘le NHOQHWH‘HNUI cow-slam“) [O N H V“ 29 Blahver, Hveravellir,west-centra1 Iceland ______________________ do 9 30 Crystal Hot Springs, Box Elder County, Utah ____________ D. E. White, in this paper, table3 _______ 11 31 Stinking S rings, 6.8 miles northwest of Corinne, Utah. White, Hem, and Waring (1963, table 15)_ 5 32 Utah Hot prings, Weber-Box Elder County line, Utah.-- - 56 7 This paper, table 3 ______________________ 12 33 Becks Hot Springs, Salt Lake County, Utah _____________ -___ 18 34 _____ do ---------------------------------------- 55 5 19 35 Hooper Hot Spring, Davis County 14 36 El Monte (Ogden) Hot Springs, Weber County, Utah ........... 13 37 Joseph Hot Spring, Sevler County, Uta h ................... 15 38 Red Hill Hot Spring, Sevier County, Utah ........ 16 39 Abraham Hot Springs, Juab County, Utah-__ __________ 17 40 Roosevelt Hot Spring, Beaver County, Utah- 10 41 Burgin mine, Utah County, Utah ______________________________ 4 42 ----- do ----- _ - - _ _ - ------ 7 43 ----- do ----- __ __________ 3 44 ----- do ________________________________ 5 Lake, however, is extremely small (< 0.0001), and presumably the BrzCl ratio for a salt in equilibrium with it would be at least an order of magnitude less. Fresh water dissolving such halite certainly would also have a similar bromine to chlorine ratio, and the sharp difference between Br:Cl ratios in the two types of water (as shown in fig. 3) is striking confirmation of this fact. Inasmuch as the BrZCl ratio of the Burgin mine water is approximately 0.001, the ratio is in har- mony with our conclusions reached from study of the D:H isotopes, and we conclude also that the Burgin brine is neither a diluted marine connate water nor yet a diluted inland connate water of the type represented by Great Salt Lake. White (1957), among others, stressed the high content of silica characteristic of waters from geysers and boiling springs associated with volcanic areas and, conversely, the low content of silica in connate or meteoric water. The silica content of meteoric water, however, has a substantial range, as does the silica content of thermal- spring water that issues some distance from shallow sources of volcanic heat. We have found that the ratio of silica to total solids in saline waters is appreciably different in three general groups: (1) water associated with geysers and boiling springs, (2) thermal water associated with volcanic areas but reaching the surface at temperatures less than 85°C, and (3) thermal or cool chloride water of meteoric——or connate—origin. Ratios for such waters are shown on the graph (fig. 3), which reveals that connate water and chloride water of mete- oric origin have extremely low ratios of silica to total solids; the boiling springs and geyser waters have the highest ratios of silica to solids, for they are relatively dilute, but their silica content is mainly a temperature~ dependent phenomenon caused by saturation or super- saturation of silica in very hot water. The waters from thermal springs associated with volcanism but issuing at temperatures well below the boiling point have ratios of silica to total solids that overlap the range of ratios shown by boiling springs, F24 and although their average is much below that of the boiling springs, their ratios are an order of magnitude greater than those of connate brines and brines formed by meteoric water in contact with salt beds. The Bur— gin mine water and most of the hot springs of Utah have ratios of silica to total solids that are clearly in the intermediate range, and even the lowest ratios among them are well above the ratios in connate brines or brines of meteoric origin. It might be argued, however, that the silica content in the Burgin water—about 48 ppm—suggests a meteoric origin; nevertheless, dilution of hot siliceous brines of volcanic origin by meteoric water having the composition of local meteoric water (table 3, anal. 1) could yield a brine with approximately the same silica content as the cool diluent if the amount present represented saturation at the cooler tem- perature. The siliceous chloride waters of many boiling springs and geysers have several hundred parts per million silica in solution where they first appear at the surface orifice, but the waters precipitate some of their silica load in the outlet close by and soon attain a content appropriate to the equilibrium between opal or amor- phous silica and hot water. As is well known, amor— phous silica gel is much more soluble than quartz. According to Morey, Fournier, and Rowe (1962) the solubility of quartz in pure water is very low at room temperature and rises appreciably With increasing tem- perature: 6 ppm at 25°C, 19 ppm at 63°C, 45 ppm at 94°C, 57 ppm at 100°C, and 190 ppm at 180°C; in one experiment these workers/found unequivocal evidence that quartz was deposited from a dilute but super saturated solution (80 ppm) at 25°C after 1 year. The solubility of amorphous silica is much greater, and according to these workers is about 135 ppm at 25°C and about 380 ppm at 100°C. The effect of a large content of chloride ions on the solubility and rate of reaction of quartz may be great, however, and recent work by Van Lier and others (1960) is relevant. They reported that quartz has a well-defined solubility in water at temperatures below 100°C, that the solubility ranges from about 30 ppm at 60°C to about 75 ppm at 100°C, and furthermore that sodium chloride solutions greatly accelerate the rates of dissolution and of crystallization of quartz. Their experiments showed that equilibrium was attained in a few days in chloride solutions Whereas in pure water several weeks were required. Although they concluded that the solubility of quartz in pure water is almost the same as in solutions containing less than 0.1 N sodium chloride (<5,850 ppm NaCl), they found some evidence of increased solubility of quartz in salt solu— tions of higher concentration. Inasmuch as the hot chloride water of the Burgin mine approximates 0.1N SHORTER CONTRIBUTIONS T0 GENERAL GEOLOGY sodium chloride when collected, it is probable that they have an even higher chloride content at greater depth and that the temperatures are also higher than those measured. The value of 45 ppm of silica found in this mine water corresponds with the solubility of quartz either at 94°C (201°F) according to Morey and others (1962) or at 80°C (17 6°F) according to the data graphed in figure 6 of Van Lier and others (1960). It thus seems that the amount of silica in solution in the hot sodium chloride water of the Bergin mine is appro- priate to a chloride solution that had attained equilib- rium with quartz at temperatures near the boiling point. The solutions presumably pass through the thick basement of quartzite on their way to ground- water level, and equilibrium with quartz should be effected quickly as a result of the high salt content of the water. Silica gel and opal precipitate on hot—spring aprons, but quartz apparently does not; it is rarely deposited from solution in the laboratory below 100°C. Quartz readily precipitates from solution at elevated tempera— tures, however, and if hot siliceous chloride waters deposited quartz as the stable phase at depth, the silica content of the solution would drop to a fairly low figure. Such a mechanism may explain the low silica content (72 ppm) of spring 21 of the Steamboat Springs area, noted by White, Brannock, and Murata (1956, p. 49). Still another mechanism would give the low ratio of silica to total solids: If magmatic vapor rose into an evaporite series and there condensed on contact With either connate water or brine derived from meteoric water, the salt content might increase relative to the magmatic contribution while the silica content de- creased. It may be significant, however, that, although many of the Utah hot springs having low ratios of silica to total solids are associated with modern evaporite basins, salt deposits have not yet been reported from the nearby Tertiary and pre—Bonneville Pleistocene deposits. The K:N a ratio is also helpful in distinguishing between waters of meteoric origin and those of volcanic origin. As shown on the graph (fig. 3), the KzNa ratios in brines of meteoric origin range from 0.022 to 0.045 and average about 0.03; in contrast, ratios in water from volcanic sources are much higher, ranging from 0.04 to 0.25. The average of the K:N a ratios of hot springs associated with volcanic activity is 0.10, which is almost exactly the KzNa ratio of the hot saline Burgin mine water. This ratio is so far above that of the average brine formed by meteoric water (0.03) that it alone would strongly suggest a volcanic origin for the water. The relatively high KzNa ratios of boiling or hot saline springs in volcanic areas has l l . UNDERGROUND TEMPERATURES AND EAT FLOW, EAST TINTIC DISTRICT, UTAH often been ascribed to a temperature dependence—th hotter the water the greater the proportion of potassiu . No such relation is evident in the family of Utah hot springs to which the Burgin water belongs. As show: in table 3, the KzNa ratios of eight springs havin temperatures of 130 to 139°F range from 0.05 to 0.2 , of three having temperatures of 140 to 149°F range fro 0.03 to 0.09, and of two at temperatures at 169°F and 179°F are 0.12 and 0.07. No relation of KzNa ratios to temperature are discerned. Finally, because of its twofold significance, the ratio of Li:Na should be considered. White (1957) has stressed the fact that, relative to other components, lithium is higher in thermal chloride water of volcanic association than in any other known type of natural water; he suggests that lithium and other alkali metals were transported as soluble alkali halides in a dense vapor phase from a magma to the base of the zone saturated by ground water, where the vapor condensed to form a brine unusually high in lithium. The ratios of Li:Cl shown in table 8 bear out his observation that thermal chloride waters associated with volcanic areas have a much higher Li:Na ratio than do brines formed by meteoric water in contact with salt. The Li:Na ratio of the Burgin chloride water fits that of the group of hot chloride waters of volcanic association and is thirteen times greater than that of the average of con- nate water and eighteen times greater than that of meteoric water in contact with salt deposits. TABLE 8.—Lithium to sodium ratios of various chloride-type waters [Ratios computed from analyses used for fig. 3, in which lithium is reported] Ratio of Li:Na Type of chloride water Average Median ‘ Range Geyser-s and boiling springs ___________________ 0. 00940 0. 00900 0. 00010—0. 01900 Hot springs in volcanic areas _______ ._ 00440 . 00220 . 00050— . 01100 Burgin mine, chloride waters _______ 00430 . 00400 . .00220— .00660 Hot springs in Utah, T>125°__.. ..... .00220 .00110 .00050— .01100 Connate waters ______________________________ .00033 .00025 .00011— . 00074 Meteoric waters in contact with salt __________ 00024 . 00021 .00011— . 00055 White and Brannock (1950) argued that an outstand- ing characteristic of a system of thermal springs deriving its heat from a normal temperature increase with depth is a lower gradient than that of the surrounding region. In the East Tintic thermal area Where the heat flow above the water table is abnormally high, the heat flow is also high for the few holes in which gradients have been measured below the water table. Neither is there any evidence of abnormally low heat flow in the surrounding region. In as much as hot ground water moves eastward to Utah Valley, the heat flow measured above the water table to the east must be well above average, but no data are available for deep heat flow there. .To the west in, the main Tintic and Mintintic F25 areas, the heat flow seems about average except for a few places where it is abnormally high. Although data for regional heat-flow isograms are lacking, there is no evidence that any area has given up heat at depth to deeply circulating ground water. The heat flow (ac- cording to our measurements) in the Mintintic area on - the west side of the range is greater than 2 mc cm‘2 sec—1, and the heat flow on the eastern slope in Govern— ment Canyon, 5 miles south of the map area of figure 6 is 1.9:t0.3 mc cm‘2 sec‘1 according to Roy (1963, p. 49, see footnote 1, p. F2). The heat-deficient areas, re- quired by the assumption that deep meteoric water moving laterally has abstracted heat at depth and has then risen again in the East Tintic thermal area, are at present (1964) unknown. Lacking evidence of these areas, we can only say that the three—dimensional heat- fiow pattern as currently known, as well as the com- position of the brine, argues for a volcanic source of heat. Tertiary nonmarine evaporites have been suggested . as the source of salts in the hot chloride spring waters of Utah, but the lack of any known salt-bearing evap- orites of Tertiary age anywhere in Utah argues strongly against this interpretation. No salt is known in the thick Salt Lake Formation of Tertiary age present in the valleys adjacent to the East Tintic Mountains. Furthermore, the BrzCl and Li:Na ratios of recent brines in the Great Salt Lake Desert, which should simulate those in a local Tertiary closed-basin evapo- rite, suggest a difl’erent interpretation. The ratios in brines that represent meteoric water in contact with salt deposits of the desiccated Lake Bonneville, as represented by a composite of many samples taken from holes a few feet deep in the Great Salt Lake Desert are totally unlike the ratios in the Burgin water (fig. 3 N0. 3). Meteoric water that has passed through either marine or continental evaporites seems unlikely to have the composition or the temperature required to explain the hot chloride water of the Burgin mine. If Burgin water is a composite of magmatic water and deep meteoric water that circulated through pre-Tertiary evaporite beds, the meteoric water must have picked up its addi- tional load of salt from evaporites far below the sur- face; yet no such evaporites are known near the East Tintic district. Evaporites of Jurassic age are present some 25 miles to the southeast and may lie below another thrust plate well below the thick Paleozoic sec- tion in the footwall of the East Tintic thrust. If so, meteoric water moving down to great depths, passing through the evaporites, mixing with magmatic emana- tions, as at Larderello, Italy (Elizondo, 1964; Facca and Tonani, 1964), and rising through thousands of feet of sedimentary rocks in two thrust plates might even- F26 tually issue as a hot saline spring in the East Tintic thermal area. This possibility is assuredly worthy of consideration and is discussed further in the following pages. The best alternative to this explanation as- sumes a magmatic emanation which follows deep frac— tures and heats the ground water to the temperatures observed. Geophysical work or deep drilling may solve the problem of the origin of these thermal springs more satisfactorily, but at the present time we favor the hypothesis that both the heat and mineral content of the Burgin brine are chiefly of volcanic origin. LATITE RIDGE THERMAL AREA South of the Apex Standard thermal trough the temperature at elevation 4,550 feet increases steadily through a distance of 2,000 feet to the southeast at a rate of 25°F per 100 feet horizontally; it reaches 163°F at drill hole EP—2, the southeasternmost drill hole for which thermal data are available. If this lateral gra- dient persists for another 1,600 feet, the temperature would reach the boiling point of water at the elevation of the ground—water surface. Such temperatures suggest that at depths of a few thousand feet temperatures may be in the range of geothermal power requirements. Although high temperature is the first essential of geothermal power, it is becoming increasingly evident that several other factors are significant. Ideally a geothermal power supply is the geologic analog of a steam dome on a boiler which has an ample supply of water and heat. Such a combination may not be as rare as it might at first seem. The geologic equivalent of the steam dome and boiler does in fact exist at Larde- rello, Italy. There, drill holes about 2,000 feet deep tap steam that produces the cheapest electrical power now available in the world—about 2.4 billion kilowatt hours per year from generating plants with a total installed capacity of 300,000 kw (Electrical World, 1963). The cost is 2.55 mills per kilowatt hour for the most modern plants (Facca and Ten Dam, 1963, p. 12). The steam wells are used for about 20 years before the accumula- tion of material deposited from the steam makes it necessary to drill another hole. This area has no known volcanic activity other than the steam field (Bozza, 1961). A review of the geology of the Larderello field is instructive. Steam at high pressures and temperatures (200° to 250°C according to Facca and Ten Dam, 1963) is found in a somewhat broken horstlike fault block that is structurally high; the steam reservoir is in gypsiferous Jurassic evaporites under the Argille seag- liose, a thrust plate of impervious broken claystone, serpentinized rock, limestone, and sandstone. For 6,000 feet below the thrust plate, convection of hot SHORTER CONTRIBUTIONS TO GENERAL GEOLOGY saline ground water maintains nearly constant tem- peratures as the water carries heat upward from its deep and presumably volcanic source. The broken horst simulates a regional steam dome, and to a geolo- gist the analogy between the steam trap and a geologic trap for natural gas or petroleum is striking. Oil seeps or gas leaks (as at Baku, U.S.S.R.) are common over many gas and oil traps and have led to discovery by drilling at the site of the seep. Usually the seeps are found eccentric to the productive struc- ture, however, and the position of the trap must be found by geological and geophysical studies. At Larderello the “geothermal seeps” of steam and hot springs occur at the surface in the broken horst where a few fairly tight fractures in the confining thrust plate permit the slow escape of the steam from the cavernous evaporite beds below. The analogy of steam trap and gas trap is apparent. Geothermal exploration throughout the world has been confined almost solely to areas of boiling springs and geysers, but as the geologic factors that favor geothermal power production become better under- stood, hot springs probably will be regarded more and more as “geothermal seeps” which will serve as the locus of geologic studies and geothermal surveys that use techniques of the sort described earlier in this report (but see also Facca and Ten Dam, 1963). The possibility that the Latite Ridge area has geo- thermal power potential should be given consideration. Additional drill holes in the area to the southeast are needed, but the geothermal gradient of about 10°F per 100 feet in EP—2 demonstrates the presence of a heat source worthy of consideration, if a favorable structure is present. The region southeast of the Apex Standard thermal trough may have some similarity to the Larderello structure and its geology is therefore of considerable interest. Much of the East Tintic Mountains south of the Hansen fault is covered by lava, but the drill holes southeast of the Apex Standard mine and a few windows through the lava on the western side of the range show the presence of beds low in the stratigraphic sequence and indicate that the area is geologically high. The western side of the range is bordered by a Basin and Range fault which probably has a displace— ment of about 6,000 feet, the western or basinward side being faulted down. At the southern end of the mountains the beds dip south or southeastward and the lower part of the Paleozoic section is exposed near the southern edge of the lava field. Here, too evidence is persuasive that the bedrock is structurally high. It is very probable that all these beds are in the upper plate of a major thrust fault, but the position of the thrust beneath the UNDERGROUND TEMPERATURES AND HEAT‘ FLOW, EAST TINTIC DISTRICT, UTAH lavas on the eastern side of the East Tintic Mountains is uncertain. The northern limit of this structurally high unit is probably a northeasterly shear zone that is parallel to the major shear faults of the East Tintic Mountains. Pyritic alteration in the lava southeast of the Inez and Hansen faults suggests that several northeasterly fractures underlie the lava within a few thousand feet of the Inez fault. North of the Hansen fault, beds of middle and late Paleozoic age are present in the foot- wall of the East Tintic thrust, which apparently is cut off by the northeastward-trending Inez shear zone. To the southeast, however, Tintic Quartzite or the immediately overlying Ophir Formation lies directly below the lava. A northeastward—trending shear zone that separates Pennsylvanian rocks on the northwest from Cambrian rocks on the southeast cuts through West Mountain at the southeastern side of Lake Utah 10 miles to the northeast of the place where the Inez fault would reach the eastern edge of the East Tintic range. This major shear zone is believed to be related to the Inez fault and it probably represents a persistent regional tear fault (Morris and Shepard, 1964). Hole ET—89 (pl. 3, co- ordinates 29,245 N., 29,650 E.), which is southeast of the Inez fault line, cuts quartzite below the lava and cuts a sandstone and shaly dolomite of unknown age below a braccia zone that may represent the footwall of a thrust lying below the footwall of the East Tintic thrust to the north. Although the Inez fault may be a tear fault affecting only a warped upper plate of the East Tintic thrust, the logs of diamond—drill holes suggest that it is a major shear zone that cuts deeper and reaches a lower thrust fault of large displacement. Southwest of the Apex Standard mine there are many minor shear zones and faults; as a result, progressively older formations crop out to the southeast. The drill logs and surface exposures clearly show that the northwest sides of the Hansen and Inez faults are downdropped relative to the southeast sides. The evidence available at this time (1964) indicates that the northeasterly zone of shearing along the Inez fault has affected the footwall as well as the hanging wall of the East Tintic thrust fault. If the sandstone, shale, and dolomite found below the quartzite in drill hole ET 89 were present at depth southeast of hole EP 2 and had an average thermal conductivity of 6.0X10‘3, the temperature would increase vertically at a rate of about 6.5 °F per 100 feet. This gradient, if maintained, would give a temperature of approximately 460 T (242 °C) at sea-level elevation under drill hole EP 2; saturated steam at this tempera- ture has a pressure of 450 pounds per square inch. The gradient, however, would not persist where convection F27 behame a major factor, as beneath a thrust plate that covered Jurassic evaporites. Inasmuch as the gradient seems to be increasing to the southeast, higher tempera— tures should be found at shallower depths. The evidence now available indicates that the terrane underlying the lava in the East Tintic Mountains southeast of the Inez fault zone is structurally high compared with that to the northwest. The great thrust fault which underlies the southern Wasatch Mountains has brought Cambrian to Pennsylvanian rocks over gypsiferous Jurassic rocks; if this fault underlies the East Tintic Mountains also, it is very possible that the upfaulted blocks southeast of the Inez and Hansen faults are underlain by this thrust fault at a depth accessible to the drill. Surveys of the type described in this report and in the paper by Facca and Ten Dam (1963) are fairly cheap and could be used to evaluate quickly the potential of possible geothermal areas. The marked thermal gradient southeast of the Inez fault and the possibility of structures favorable for development of geothermal power in the uplifted area to the southeast should offer encouragement for further exploration and geothermal surveys in that area. REFERENCES CITED Birch, A. F., and Clark, Harry, 1940, The thermal conductivity of rocks and its dependence upon temperature and composi- tion: Am. Jour. Sci., v. 238, pt. 1 of no. 8, p. 529—558; pt. 2 of no. 9, p. 613—635. Bozza, J ., 1961, Sul origine de la termalita nelle aeque e nel vapore endogeno: Larderello, Italy, Geothermal Co. (print- ed in Italian, English, and French). Croft, A. R., and Monninger, L. V., 1953, Evapotranspiration and other water losses on some aspen forest types in relation to water available for stream flow: Am. Geophys. Union Trans, v. 34, no. 4, p. 563—574. Electrical World, 1963, New geysers to harness: Electrical World, v. 160, no. 24, p. 15-18. Elizondo, J. R., 1964, Prospection of geothermal fields and in- vestigations necessary to evaluate their capacity, in v. 2, Geothermal energy, [pt.] I, of Proceedings of the United Nations conference on new sources of energy, Rome, 1961: New York, United Nations, p. 3—23. Facca, Giancarlo, and Ten Dam, A., 1963, Geothermal power economics: Rome, Italy, Consiglio Nazionale Delle Ricerche, Commissione Geotermica Italiana, 37 p. Facca, Giancarlo, and Tonani, Franco, 1964, Natural steam geology and geochemistry, in v. 2, Geothermal energy, [pt.] I, of Proceedings of the United Nations conference on new sources of energy, Rome, 1961: New York, United Nations, p. 219-229. Gates, J. S., 1963, Hydrogeology of Middle Canyon, Oquirrh Mountains, Tooele County, Utah: US. Geol. Survey Water~Supply Paper 1619-K, p. K1—K40. Gott, G. B., and Schnabel, R. W., 1963, Geology of the Edge- mont NE quadrangle, Fall River and Custer Counties, South Dakota: US Geol. Survey Bull. 1063—E, p. 191—215. F28 Hales, A. L., 1937, Convection currents in geysers: Royal Astron. Soc. Monthly Notices, Geophys. Supp., v. 4, no. 1, p. 122—131. J aeger, J. C., and Le Marne, A., 1963, The penetration of ventila- tion cooling around mine openings and extrapolation to virgin rock temperatures: Australian Jour. Applied Sci., v. 14, no. 2, p. 95—108. Junge, C. E., 1963, Air chemistry and radioactivity, V. 4 of International geophysical series: New York and London, Academic Press, 270 p. Karplus, W. J ., 1958, Analog simulation: New York, McGraw Hill Book Co., 434 p. Y Lachenbruch, A. H., and Brewer, M. C., 1959, Dissipation of the temperature eflect of drilling a well in Arctic Alaska: U.S. Geol. Survey Bull. 1083—C, p. 73—109. Lindgren, Waldemar, and Loughlin, G. F., 1919, Geology and ore deposits of the Tintic mining district, Utah: U.S. Geol. Survey Prof. Paper 107, 282 p. Lovering, T. S., 1948, Geothermal gradients, recent climatic changes, and rate of sulfide oxidation in the San Manuel district, Arizona: Econ. Geology, v. 43, no. 1, p. 1—20. Lovering, T. S., and others, 1960, Geologic and alteration maps of East Tintic district, Utah: U.S. Geol. Survey Mineral Inv. Map MF—230. Lovering, T. S., and Goode, H. D., 1963, Measuring geothermal gradients in drill holes less than 60 feet deep, East Tintic district, Utah: U.S. Geol. Survey Bull. 1172, 48 p. Morey, G. W., Fournier, R. 0., and Rowe, J. J., 1962, The solubility of quartz in water in the temperature interval from 25° to 300°C: Geochim. et Cosmochim. Acta, v. 26, p. 1029-1043. Morris, H. T., 1964, Geology of the Eureka quadrangle, Utah and Juab Counties, Utah: U.S. Geol. Survey Bull. 1142—K, 29 p. Morris, H. T., and Shepard, W. M., 1964, Evidence for a con- cealed tear fault with large displacement in the central East Tintic Mountains, Utah, in Geological Survey research, 1964: U.S. Geol. Survey Prof. Paper 501—0, p. 019—021. O SHORTER CONTRIBUTIONS T0. GENERAL GEOLOGY Richardson, G. B., 1907, Underground waters in Sanpete and central Sevier valleys, Utah: U.S. Geol. Survey Water- Supply Paper 199, 63 p. Spilhaus, A. F., 1938, A bathythermograph: Jour. Marine Re- search, v. 1, no. 2, p. 95—100. Stearns, N. D., Stearns, H. T., and Waring, G. A., 1937, Thermal springs in the United States: U.S. Geol. Survey Water— Supply Paper 679—B, p. 59—206. U.S. Weather Bureau, 1952, Climatological data, Utah; Annual summary, 1951: U.S. Dept. Commerce, v. 53, no. 13, p. 189. Valyashko, M. G., 1959, I‘eongnH Bpoma B nponeccax rano- renesa n ECHOJIBBOBaHKe conepmannx BpOMa B Kaqecrne rene'mqecrcoro n IIOKOKOBOI‘O rcpn'repna [Geochemistry of bromine in the processes of salt deposition and the use of the bromine content as a genetic and prospecting criterion. With English abs]: Internat. Geol. Cong, 20th, Mexico City, 1956, Symposium de exploracion geoquimica, v. 2, p. 261—281. See also translation from Geokhimiya, 1956, published in English in Geochemistry, 1960, v. 6, p. 570—589. Van Lier, J. A., de Bruyn, P. L., Overbeek, J. Th. G., 1960, The solubility of quartz: Jour. Phys. Chem., v. 64, p. 1675—1682. Van Orstrand, C. E., 1924, Apparatus for the measurements of temperatures in deep wells by means of maximum thermom- eters: Econ. Geology, v. 19, no. 3, p. 228—248. White, D. E., 1957, Magmatic, connate, and metamorphic waters: Geol. Soc. America Bull., v. 68, no. 12, pt. 1, p. 1659-1682. White, D. E., and Brannock, W. W., 1950, The sources of heat and water supply of themal springs, with particular reference to Steamboat Springs, Nevada: Am. Geophys. Union Trans, v. 31, no. 4, p. 566—574. ' White, D. E., Brannock, W. W., and Murata, K. J., 1956, Silica in hot spring waters: Geochim. et Cosmochim. Acta, v. 10, nos. 1—2, p. 27—59. White, D. E., Hem, J. D., and Waring, G. A., 1963, Chemical composition of subsurface waters: U.S. Geol. Survey Prof. Paper 440—F, 67 p. PROFESSIONAL PAPER 504-1: UNITED STATES DEPARTMENT OF THE INTERIOR PLATE 1 GEOLOGICAL SURVEY 18000 E 20000 E 22000 E g: 1 . 26 000 E 28000 E 3}) 0°24%00 N 44 000 N; f. «gm arc/31%“ s‘ “’72 ‘7 MAY 4 1956 ‘2 <15 4’”! ME ME" 42 000 N EXPLANATION X “t Tertiary lava Paleozoic undivi 40 000 N 40000 N Cambrian Tintic Quartzite Contact Steepkfaults and shear zones N »»\\~«.\\\\\Mmuwfik\\«%ve\\v delineate Wisely/Maw W. MAW Thrust fault Sawteeth on upper plate OGS 37 US. Geological Survey drill hole less than 60 feet deep .30 5 Drill hole more than 100 feet deep BC, Bear Creek CC, Chief Consolidated EP, Eureka Prospect GS, US. Geological Survey G8 7 t, US. Geological Survey Trixie N, Newmont N L, North Lily TS, Tintic Standard 4595 LEVEL 1 5104 I__I_09__‘(__ 38000 N 38000 N Mine workings Showing approximate elevation, in feet. Location of measured temperature shown by dot; dots not on illustrated workings are points in raises or other workings not shown -— ————— —o 36 000 N Flat or inclined drill hole Inclined Winze 54.4; gr 4.6 q 4.0; W 104 4771; 93.6 Thermal data 54.4, mean annual surface temperature, in degrees Fahrenheit. Parentheses indicate figure is only approximate and is from surface isothermal map gr 4.6, geothermal gradient, in degrees Fahrenheit per 100 feet q 4.0, heat loss in microcalories per square centimeter per second, at the surface w 104, calculated temperature, in degrees Fahrenheit, at ground-water level (elevation about 4,550 feet) 4771, elevation, in feet, where temperature was meas- ured closest to ground-water level 34000 N 93.6, temperature, in degrees Fahrenheit, measured closest to ground-water level Note: Not all data may be shown in all boxes , 55 54 Mean-annual-temperature isotherms at surface Isotherm interval 1°F. Isotherms in areas of sparse data drawn to show similar relation to topography observed in areas of adequate data 32 000 N 30 000 N 28 000 N 26000 N \ *7 31’“? ‘ ~~ fl, V ' I, i, _ i: I . , ' ,. ‘ .l 26000 N 18000 E g"- ; 24000 E 26000 E 28 000 E . ,, ‘ 24 000 N INTERIOR—GEOLOGICAL SURVEY, WASHINGTON. D. C.—1965—665028 30 000 E Base from U.S. Geological Survey topographic quadrangle: ° SCALE 1:9600 Geology generalized from T. S. Lovering and others (1960) Eureka, 1:24,000, 1954 24 000 N , 1/2 1 MILE l‘—*-—l l TRUE NORTH .5 1 KILOM ETER APPROXlMA‘lE MEAN ‘ ' *i Tl DECLlNATION,1965 I UTAH A RANGLE CONTOUR INTERVAL 25 FEET EASTHNW THERMAL DATUM IS MEAN SEA LEVEL QUADRANGLE LOCATION AREA LOCATlON MAP OF EAST TINTIC THERMAL AREA, UTAH, SHOWING ISOTHERMS REPRESENTING MEAN ANNUAL TEMPERATURE AT THE SURFACE, MAJOR GEOLOGIC FEATURES AT THE WATER TABLE (ELEVATION APPROXIMATELY 4,550 FEET), AND SURFACE AND SUBSURFACE THERMAL DATA UNITED STATES DEPARTMENT OF THE INTERIOR GEOLOGICAL SURVEY ITEMPERATUREfIN DEGREES FAHRENHEIT 70 80 90 PROFESSIONAL PAPER 504~F PLATE 2 STRUCTURE, ALTERATION, AND ROCK FORMATION I 7/52 8/52 59.1 I 59.4 TOP OF PERCED WATER TABLE 1‘ =5; k=4.9 1 2 12/47 3/52 64.0 64.5 - .v BOTTOM OF PERCHD WATER TABLE i F =2.58; k=9.3 1 I‘ =5.354; k=4.50 (laboratory value) Pyritic alteration Drilling history Hole churn drilled from surface (elevation 5,636 ft) to depth of 1100 feet between July 21 and September 13, 1947 (52 days). Hole deepened with core drill from 1100 to 1600 feet between May 31. and August 1. 1950 (63 days), last 100 feet drilled in 3 days; hole closed at 1510 feet immediately. ‘ Temperature at depth of 1500 feet measured on August 3, 4, 5. 6, and 7, 1950 LEVEL OF-.,‘GROUND WATE‘R' ”4'546 FEET I‘= 6.43; k: 3.75 1 Tertiary Packard Quartz Latite Moderate to weak argillic I‘=5.66; k=4.31 Transition Argillic and calcitic F: 14.8; I<=1.61 Basal Calcutic tuff F=3.62; k=6.71 Devonian, Silurian, Ordovician Bluebell Formation Dolomite and minor shale Fault breccia F=2.85;k=8.513 Dolomite and dolomitic limestone F=3.78; k=6.41 Quartz veinlets . Temperature measured with a maximum thermometer 1 Temperature measured with a resistance thermometer -)(- Temperature measured with a bathythermograph Solid lines approximate true temperature gradient Broken lines show gradients distorted by drilling 11: calculated from L” F" I‘ kw:4.50 (Table Lusing laboratory value It for Tpr, p) I‘W=5.355 (this graph) 2Laboratory value is 4.93 I EXPLANATION I, ‘. 1’, "h V 3Laboratory value of dry hydrothermal dolon‘iite' is 10.63; laboratory value of dry limestone (Teutonic Limestohelis 656 ‘Laboratory value of wet lwdrothetmal dolomite is 1185 I‘ Geothermalgradient,in degrees Fahrenheit perlOO feet .‘8/50 ' I 107.5, Time and temperature data Upper Iine gives monlh and‘lyearof measurement; Iower Iine degrees,_Fahrenheif 12/47 k , Conductivityjn cgs units times 103 V (Bottom—hole temperature) Probable gradient " I‘ =2.02; k=11.904 lnterbedded dolomite and sandstone; minor breccia zones Dolomite and minor breccia zones Devonian Victoria Formation Sandy dolomite dolomite, and jasperoid Leached dolomite, Limy dolomite and dolomite LITHOLOGY, TEMPERATURES, AND GRADIENTS IN DRILL HOLE GS IIN EAST TINTIC DISTRICT, UTAH, AT VARIOUS TIMES, AND CONDUCTIVITIES CALCULATED FROM GRADIENTS 776-666 0 165 (In pocket) UNITED STATES DEPARTMENT OF THE INTERIOR GEOLOGICAL SURVEY : 2;] 18 000 E 20000 E 44 000 N ._ . . . {I /3 ‘ . a I, a a 5“ \‘xK 3}; W \ x... 0550mm mm ffwv v1]. {’1- ,/ / / WWW—s...“ \ / r $7 \N, ‘M /. 42 000 N 40 000 N 38000 N ~ ' 36000 N 28 000 N 26 000 N 24 000 N " ‘ :24» 18000 E Base from U.S. Geological Survey topographic quadrangle: Eureka, 1:24,000, 1954 APPROXIMATE MEAN DECLINATION,1965 24 000 E 22000 E {,ng WI 24000 E SCALE 1:9600 1/2 .5 CONTOUR INTERVAL 25 FEET DATUM IS MEAN SEA LEVEL 1 KILOMETER I 30 000 E 28 000 E 44000 N r 42 000 N 40 000 N or 38000 N 36 000 N 34 000 N 32 000 N 30 000 N 28 000 N ‘ 26 000 N 24 000 N 28000 E 30 000 E INTERloR—GEOLOGICAL SURVEY. WASHINGTON. D C,—1965~665028 Geology generalized from T. S. Lovering and others (1960) I UTAH RANGLE EAST TINTIC THERMAL QUADRANGLE LOCATION AREA LOCATION MAP OF EAST TINTIC THERMAL AREA, UTAH, SHOWING ISOTHERMS AND MAJOR GEOLOGIC FEATURES AT THE WATER TABLE (ELEVATION APPROXIMATELY 4,550 FEET), AND SURFACE AND SUBSURFACE THERMAL DATA PROFESSIONAL PAPER 5041—13 PLATE 5 EXPLANATION ~‘ x a \\ Tertiary lava Paleozoic undivi Cambrian Tintic Quartzite Contact Steep faults andlshear zones awamamxen saw/same. N t. Thrust fault Sawteeth on upper plate OGS 37 US. Geological Survey drill hole less than 60 feet deep .80 5 Drill hole more than 100 feet deep BC, Bear Creek CC, Chief Consolidated EP, Eureka Prospect GS, US. Geological Survey G3 7 t, US. Geological Survey Trixie N, Newmont N L, North Lily TS, Tintic Standard 4595 LEVEL 1 5104 131:: Mine workings Showing approximate elevation, in feet. Location of measured temperature shown by dot; dots not on illustrated workings are points in raises or other workings not shown ______ _. Flat or inclined drill hole Inclined winze 54.4; gr 4.6 a 4.0; W 104 4771; 93.6 Thermal data 54.4, mean annual surface temperature, in degrees Fahrenheit. Parentheses indicate figure is only approximate and is from surface isothermal map gr 4.6, geothermal gradient, in degrees Fahrenheit per 100 feet q 4.0, heat loss in microcalories per square centimeter per second, at the surface w 104, calculated temperature, in degrees Fahrenheit, at ground-water level (elevation about 4,550feet) 4771, elevation, in feet, where temperature was meas- ured closest to ground—water level 93.6, temperature, in degrees Fahrenheit, measured closest to ground-water level Note: Not all data may be shown in all boxes I40 Isotherms showing temperature (°F) at water table Isotherm interval 2°F - 4 "o W 0%)0 1a ”9% UNITED STATES DEPARTMENT OF THE INTERIOR ' 09‘A PROFESSIONAL PAPER 504—17 ‘5 GEOLOGICAL SURVEY _, PLATE 4 ”b 112°07’30” R. 3 w. R. 2 w. / (ALLENS RANCH) 40°00/ _ _ a _ _ _ , -- _ . __ . . If “\ 112000 40°00 __ No, _. K5. ' . T 10 s. EXPLANATION Quaternary surficial deposits Tertiary intrusive rocks Tertiary lava Paleozoic rocks If; 5730” Contact Dashed where approximately located l! l Jl m m Area where some precipitation is assumed to be available for recharge of ground water in East Tintic thermal area north of Apex Standard ther— mal trough;western limit is hydrologic divide, east- ern limit is east edge of area studied, northern and southern limits are defined by major faults o 4797 Elevation of ground water in mine or drill hole 6000 6/00 Hydrologic contours on deep ground-water surface Contour interval 100 feet (SANTAQUIN I: 62 500) 55’ /'\\\\:7 "’ , . . . ‘ , ‘\‘ \, , .1 , .1 ~ , \i . . -/r l . 4 A ,_. (I .3 1.x 2 ; .I M I ,1 , ./~ ~ ‘ Y, .~r,I - ‘1 ._ 4 , \. _ I , .j , \ ‘ I < c ,, " / , I \ ' ‘, .» .\ , X \' 4, I ‘ < “<5 _, ‘ ’ ’ ‘ , \ . " ’ ~‘ _«, B: KI “» ' 3 " “ ' W _, ‘, I, M, M, . \ : _ . . / '\¢_, .. ,. - ,, » ‘w _{ , ,t i .y m / I /"//‘.‘,- ’_\ ,. I 3 , ”I ,2. . 4, . ‘ _ . E \ . x , . \I ‘ f ‘ .. ., . p. .. , <- I. I ' I“ ~ ‘ \< , ‘ ‘lrl‘ I \w ‘ . ‘ ’ H , /- . 4 ,\ : ‘~ ’ . $0 ~ " , I ‘1‘, ' » . ,,;, «v- .A » » ‘ ~ , ' , ! ~ . . I v~ . , ‘_ < ‘ I / "m -, « . \ .5 J ~, __ - ,. .M/ ,1 \ ,. 1x \ . , r. «I .w i /. , ._ . I I , , , , , ' , < I III I I" l . N ’y \‘I' :‘ \‘ \/‘ ‘ \I‘ I l‘: M ~ " ;‘ » ‘ ‘ ~ 1/ «\ , , ' ; L, ” i U« A , T i K ‘- F ' ‘ Y " , ' In. . w , I ”I _.. , .4 6>OKO .‘I . T. 105. T. II S. \ \. .\ VIM? I: OIJIINIIIIIII it - I \\ / N?“ L ml \\\ $ \ \. IIIIIL‘I \\ M II __ r .. a M , \ n .w‘w VJ » \\ \ I /"—" 4, _ — if . I WvQJO/kjrf‘ . ‘ ‘ “I ‘ A , \\ LOW/KM I _. H MVAA.A3\JL~ \‘P‘\.\.b{ ,n/ \\\\ .I Q I z 5.33 L: Jk "O V“- Inky/R v f ,. \ » r I. L_.L \4 -41th—gz—fi ' l V Qlil/ \\ ‘»;,—- H , J I, “1 i. \\ 3905230» " ,, » _ ;.‘. . 41? '7 -y " c 3ng% ”WM/79X MIRA \ 112°07'30" R-3 W- - - ' -\_LI ll ”R. 2 W.1120 INTERIOR—GEOLOGICAL SURVEY, WASHINGTON, D, C.—1965—665028 39°52’30” 00’ Geology generalized from H. T. Morris (1964, pl. 1) 0" SCALE 1:24 000 UTAH HG "7 ‘3 o I 1 1/2 o 440 Base by U.S. Geological Survey, 1954 4 \t‘ o Iv“ TRUE NORTH APPROXlMATE MEAN DECLINATION,1965 QUADRANGLE LOCATION 1 .5 O 1 KlLOMETER CONTOUR INTERVAL 25 FEET DATUM IS MEAN SEA LEVEL MAP OF THE EUREKA QUADRANGLE SHOWING GENERALIZED SURFACE GEOLOGY, ELEVATION OF GROUND-WATER SURFACE AND AREA ASSUMED AVAILABLE FOR RECHARGE OF GROUND WATER IN EAST TINTIC THERMAL AREA, UTAH PROFESSIONAL PAPER 504—1: UNITED STATES DEPARTMENT OF THE INTERIOR PLATE 5 GEOLOGICAL SURVEY 30000 E '- 0 \r'.‘ , , , ”“9“ 44000 N112 Us 18000 E 20000 E 22000 E Rilh. 24000 E 26000 E 28000 E . , 44000 N 42 000 N 42 000 N EXPLANATION Tertiary lava Paleozoic undivided 40000 N 40 000 N Cambrian Tintic Quartzite cater Steep \faults and shear zones \- Mull ,l«\\/flw.m..\vy/re,mme>\\ mmeawu NUMVM Thrust fault Sawteeth on upper plate OGS 37 . US. Geological Survey drill hole less than 60 feet deep / ./ .Bc 5 i , .\ 38 000 N Drill hole more than 100 feet deep BC, Bear Creek CC, Chief Consolidated EP, Eureka Prospect GS, US. Geological Survey G8 7 t, US. Geological Survey Trixie N, Newmont N L, North Lily TS, Tintic Standard 4595 LEVEL 1 5104 ll_00_.<: 38000 N_ * Mine workings Showing approximate elevation, in feet. Location of measured temperature shown by dot; dots not an illustrated workings are points in raises or other workings not shown — ————— —o 36 000 N Flat or inclined drill hole Inclined winze 54.4; gr 4.6 q 4.0; W 104 4771; 93.6 Thermal data 54.4, mean annual surface temperature, in degrees Fahrenheit. Parentheses indicate figure is only approximate and is from surface isothermal map gr 4.6, geothermal gradient, in degrees Fahrenheit per 100 feet q 4.0, heat loss in microcalories per square centimeter per second, at the surface w 104, calculated temperature, in degrees Fahrenheit, 39 o 57: 30: at ground-water level (elevation about 1,,550feet) 4771, elevation, in feet, where temperature was meas- ured closest to ground-water level 34 000 N 93.6, temperature, in degrees Fahrenheit, measured closest to ground—water level Note: Not all data may be shown in all boxes 3 Isograms showing heat loss, in microcalories per square centimeter per sec (q/sec) Isogram interval 1 q/sec 32 000 N 30 000 N 6‘.1;W V 5490'51 28000 N v 28000 N 26 000 N 26 000 N 2 18 000 E » 20 000 E 22 000 E R, 2 w, 24 000 E 26 000 E 28 000 E l ’9 ”'0'?" 3‘17" 30 00024EOOO N . INTERIOR—GEOLOGICAL SURVEY, WASHINGTON, D C —1965—665028 24000 NW Base from U.S. Geological Survey topographic quadrangle: 1555“ SCALE 119600 Geology generalized from T. S. Lovering and others (1960) Eureka, 1:24,000, 1954 1 /2 TRUE NORTH I .5 1 KILOMETER UTAH §——4 — APPROXIMATE MEAN l RANGLE DECLINATION, 1965 DATUM Is MEAN SEA LEVEL QUADRANGLE LOCATION AREA LOCATION MAP OF EAST TINTIC THERMAL AREA, UTAH, SHOWING HEAT LOSS AT THE SURFACE, IN MICROCALORIES PER SQUARE CENTIMETER PER SECOND, MAJOR GEOLOGIC FEATURES AT THE WATER TABLE (ELEVATION APPROXIMATELY 4,550 FEET), AND SURFACE AND SUBSURFACE THERMAL DATA